Mission: melt shop revitalization at John Deere.

The Mission: "To develop a melt business plan to support long range business improvement," including the following goals:

* support 100,000 tons/year of quality castings;

* reduce direct costs by 25% (less charge materials);

* achieve environmental compliance;

* achieve quality improvements;

* reduce safety incident rate.

The Situation: John Deere Foundry, Waterloo, Iowa, is a captive gray and ductile iron foundry that produces components for agricultural equipment. The 286-employee foundry includes 1.2 million sq ft in buildings and is the last remaining Deere casting facility.

Since 1972, John Deere Foundry had operated electric arc furnaces (EAF) for primary melting with a capacity of 13-15 metric tons/heat each. The EAFs duplexed into six channel holding furnaces, each with a capacity nearing 85 tons of ready-to-pour molten metal.

Because of increasing competition from other foundries, the need to improve its productivity and concern about excessive emissions in light of the proposed Maximum Achievable Control Technology (MACT) standards, the company reviewed alternate processes for casting production in the mid-90s. It found that one of the highest costs of producing a casting centered in melt operations, prompting the decision to upgrade the melt shop in January 1997.

Casing the Assignment

Foundry management formed a team of Deere employees to research, plan and propose a revitalized melt operation. The task included benchmarking the industry, identifying available technology, projecting capital requirements and identifying areas of potential return on investment. The team studied the electric arc furnace (both AC and DC), the vertical shaft cupola, the coreless induction furnace (line and medium frequency), and the channel induction furnace, Energy demands, raw materials, refractory and environmental operating parameters all required evaluation and entered the equation for weighing different options prior to developing a final proposal for consideration.

Using a spreadsheet cost model, the team compared the different furnace options and also considered how the benefits of each melt system measured up to its list of "musts:"

* must exceed EPA compliance;

* must be capable of melting both gray and ductile iron;

* must maintain flexibility to deliver iron to all holding furnaces;

* must conform to existing hot metal and charge crane capacities;

* must not constrain molding unit productivity.

With this as a starting point, the project team began collecting cost approximations and other data for the spreadsheet (see sidebar, "Gathering Intelligence: Melt System Comparisons"). It determined that annual melt tonnage shouldn't guide the final solution-hourly molding requirements should dictate the need. Supporting this argument are shift configurations, anticipated casting requirements, casting yield and holding capacity.

Uninterrupted supplies of molten iron are essential to maintaining a high level of productivity within the operation. Using casting forecasts for each of the facility's molding lines, the team projected an hourly melt rate requirement for each unit. Based upon projected required tons for a given fiscal year and various combinations of molding line speed and run-time efficiency, the team constructed tables to determine molding shift configurations. These projections were approximations and indicated a tight schedule for the molding units.

Parallel to the molding shifts, the team also reviewed hourly pour rates and determined that a maximum rate of 80 tons/hr was realistic. From this, it determined that a melt rate of 65 tons/hr (an 18% increase over the EAF operations) would support pouring operations at 90% uptime rates for the molding units. With these melting capacity numbers in mind, the team focused on its options.

Planning a Course of Action

Based upon the information entered into the cost model, cupola and coreless furnaces came out as lower-cost options compared to EAF and channel induction furnaces. An EAF was an attractive option because it can melt almost anything and has low capitalization costs, but the team didn't feel it was the right choice for the foundry because of environmental concerns with dust generation and the large volume of effluent it creates. The channel induction furnace was appealing because of its energy efficiency and chemistry control options, but was not chosen because the foundry was concerned about the high cost of refractory, necessary scrap preheating and the difficulty in converting from gray to ductile base chemistries quickly.

The two lower-cost options presented appealing benefits. A cupola offers efficient production because of its physical design--it features inherent scrap preheating, has the capacity to produce large volumes in a single furnace and requires the lowest melt labor requirements. A coreless induction furnace has a low environmental impact from dust and slag generation, high electrical and thermal efficiency, and inexpensive refractory needs.

Arguments can be made in favor of the cupola; however, a major deterrent to its selection centered on a need to produce both gray and ductile iron simultaneously. This lone requirement eliminated the cupola from consideration because of the anticipated cost of two separate melting units. The project team singled out the coreless induction furnace as the best candidate for conversion of its melt department.

Output from the model indicated the medium frequency coreless induction furnace operating as a batch furnace could indeed reduce the cost of molten iron produced at John Deere Foundry. The savings comparing a medium frequency coreless induction furnace to the current EAF costs (in terms of money saved/ton shipped) are found in Table 1.

While a major increase in charge material cost is evident with the medium frequency furnace's requirement for smaller and cleaner scrap, the elimination of electrodes required by the EAF offsets this higher charge cost. Sizing and productivity of the new furnaces allows fewer furnaces to meet the tonnage requirements. This reduces the material and labor necessary to operate and maintain the equipment.

By design, the coreless is more efficient in terms of energy usage. However, the speed at which the new furnaces must operate requires an increase in the facility's demand setting, resulting in a negative impact on the energy demand in the saving stream.

Mission Launch: The Installation

The project team presented its findings to Deere and received approval to proceed in August 1998. The equipment to be installed included three 20-metric-ton steel shell coreless furnaces, each equipped with back tilting for slag removal, integral close-capture fume collection covers and refractory push out systems.

Due to a need for continued production, construction occurred over two phases. Phase I included the installation of one coreless furnace and its support equipment (ladle cars to move transfer ladles in toward the furnace for tapping, the charge pan, alloy batching system and dust collector) in an area vacated by the melt area maintenance support group (Fig. 1).

The first iron was tapped from the new coreless furnace on September 27, 1999, 13 months following approval to proceed with the project. While installing the first furnace (designated C3), all six electric arc furnaces continued to produce iron for the molding lines (Fig. 2). Once the new furnace achieved a level of confidence in sustained operation, three of the arc furnaces were shut down and removed to provide room for Phase II in the construction schedule, installation of both C-1 and C-2. Construction for Phase II began in January 2000, and the second furnace tapped its first iron in July 2000 with the third and final furnace starting up in early August.

The Operative in Action

Each of the three furnaces has its own charge conveyor, capable of holding the entire charge burden. Each charge conveyor sets atop an integral weight frame built into the melt deck that provides charge crane operators feedback as the conveyor is filled. A newly designed alloy addition system allows alloy batching from preprogrammed recipes for bath chemistries and temperatures.

The new furnaces use a computerized control system for overall melt furnace management (Fig. 3). This system is managed from a central elevated control room with line of sight monitoring of each furnace. Remote control stations adjacent to each furnace provide full local control for furnace operators on the melt deck.

The furnace installation also provides new dust and fume control via a 200,000 cfm (cu ft/min) pulse-jet collector. All three coreless furnaces share the single collector. The collector includes gathering conveyors, a pre-coat silo and a storage silo for dust and mixing equipment to process the dust prior to offsite reuse or disposal. Emission collection points are located at each furnace and include scrap charging, a close-capture hood above the furnace and a tapping hood over the furnace spout. To reduce overall requirements for the emissions system, dampers are used to direct exhaust flow where needed.

Analyzing the Case

With nearly 6000 heats produced to date, many of the expectations sought by the conversion from EAF to coreless induction have been realized. Below is a breakdown of savings, with total money saved/ton shipped presented in Table 1.

Melt Energy ($2.70 saved/ton shipped)--Current operations require the furnaces to run over a two-shift production schedule. Each furnace holds iron across an 8-hr shift to reduce thermal stress on the silica linings, which results in an inflated value for overall melting energy. Excluding support equipment such as cooling towers and the emission collector, a reduction in kWh/ton melted has been achieved--many heats require less than 500 kWh/ton. The global average to date for all heats, including sinter and cold starts, is 480 kWh/ ton. A comparable value for the electric arc furnaces averaged between 540 and 560 kWh/ton.

Energy Demand (-$3.34 saved/ton shipped)--The present production requirements and operational strategy have prevented an excess cost penalty, which eliminates a greater negative impact on the projected savings stream.

Electrodes ($33.55 saved/ton shipped)--For every ton produced with the new furnaces, the operation saves the cost of electrodes.

Charge Materials (-$20.22 saved/ton shipped)--The EAF is a forgiving furnace with the size, type and chemistry of scrap metal that is acceptable. The coreless furnace cannot accept aggressive, low cost scrap as sharp jagged edges can cut into its thinner refractory layer.

Scrap metal (such as busheling) and shredded scrap are acceptable coreless furnace material. Rust on the scrap can have a negative effect because the amount of slag generated increases as the amount of oxidation goes up. Scrap that requires more processing and must he kept under cover to prevent rust has additional costs associated with it.

Production Labor ($5.36 saved/ton shipped)--Production labor is on track with projections by the project team. The only exceptions center on stocking the alloy batching system, occasional problems with excessive slag and the intermediate refractory repairs. Overall, when compared to the resources required by the electric arc operation, manpower has remained comparable. As experience is gained and operations are tuned, an overall reduction in the production labor requirements is expected.

Refractory ($1.02 saved/ton shipped)--The learning curve for both furnace operation and refractory technique has been slow, but progressive. Bright spots in this experience include a number of campaigns in excess of 250 heats on a lining, with several exceeding 300 heats. The longest campaign to date is 427 heats.

Improvements in campaign length have been made through various changes in materials and practices. Precast shapes are being used in the tapping and slag spouts to maintain a more uniform section and reorient the interface line between the spout and main furnace lining. Changes in the sinter and cold-start procedures have accompanied experimentation with nontraditional lining materials in the upper area of the lining. An electric resistance heater was engineered and built to help in coil grout, brick and castable dry-out.

Maintenance Materials and Labor ($18.03 saved/ton shipped)--The in-house knowledge base for trouble-shooting and repair was non-existent in the beginning, and an adequate supply of spare parts was never fully anticipated. Each obstacle encountered has strengthened the maintenance and production departments' understanding of the furnaces and their operation. While costly in the short term, the experience maintenance personnel have gained will prove beneficial over the long term.

Pollution Control ($4.82 saved/ton shipped)--Beyond the reduced emissions within the building, fugitive emissions outside the building have been eliminated. The volume of dust generated by the melting process has dropped significantly. Historically, the electric arc furnaces generated between 20-40 lbs of dust/ton of molten metal produced. The coreless furnaces have generated less than 10 lbs of dust/ton of metal melted.

Waste Disposal ($0.90 saved/ton shipped)--The foundry continues to participate in a beneficial reuse program, although the volume of dust generated has decreased.

Energy Curtailment ($1.30 saved/ton shipped)--Current production schedules don't require reduced energy use.

Post-Mission Advice

Looking back, the largest problems encountered centered on the on-hand availability of spare parts and inhouse experience with this equipment. The key is to keep it common. Wherever possible, insist on part commonality with equipment and systems currently in use. From simple things like pushbuttons and indicator lights to more complex assemblies like cooling towers and hydraulic power units, the best advice for spare parts is to use components employees are familiar with.

Experience can be learned or purchased. While frustrating during the startup and debugging process, hands-on learning seemed to succeed. In hindsight, pre-startup training didn't fully prepare plant maintenance personnel for the tasks that lay ahead. Attention to details cannot be overstressed when it comes to installing a new system.

Final Analysis: Mission successful.

This article was adapted from a 2002 APS Transaction Paper (02-010).

For a free copy of this article circle No. 345 on the Reader Action Card.

Table 1

$ Saved Using Coreless Induction Instead of Electric Arc

Category                        Savings/Ton Shipped


Melt Energy                            $2.70

Energy Demand                         ($3.34)

Electrodes                            $33.55

Charge Materials                     ($20.22)

Production Labor                       $5.36

Refractory                             $1.02

Maintenance Material and Labor        $18.03

Pollution Control                      $4.82

Waste Disposal                         $0.90

Curtailment                            $1.30

Total $/Ton Shipped Savings           $44.12
Using Coreless Induction

For More Information

"Induction and Cupola Melting: A Cost Comparison Model," JR. Rotella, CMP Report No. 89-4, Center for Metals Production, Pittsburgh (1989).

"John Deere Foundry Melt Shop Revitalization Project," H. M. Ulfers, R. Christoffersen, R.S. Weaver, ISS 58th Electric Furnace Conference Proceedings (2000).

RELATED ARTICLE: Gathering Intelligence: Melt System Comparisons

When comparing potential furnace systems to implement, it is necessary to generate comprehensive data for each of the alternatives being investigated. Following is a list of considerations to take into account in the planning stages of such an operation:

Energy

* commodity and demand cost of electrical energy;

* commodity and demand cost of natural gas;

* electrode cost and supply;

* coke cost and supply.

Production supplies

* scrap cost and availability;

* alloy sizing, cost and availability.

Refractory

* type and anticipated usage;

* means of installation and labor requirements.

Manpower

* direct labor requirements;

* support systems.

Maintenance

* spare parts;

* manpower.

Emissions

* energy requirements;

* waste streams;

* means of disposal;

* beneficial reuse.

About the Authors

Stan Weaver has worked at John Deere Foundry for 25 yr in engineering and melt shop operations. He graduated from Univ. of Missouri at Rolla with an engineering degree, earned a master's degree from Univ. of Northern Iowa (UNI) in industrial technology and was the lead mechanical engineer for the melt shop revitalization project. Travis Frush has a bachelor's and a master's degree from UNI in industrial technology. He has worked at John Deere Foundry since January 2000 and is the process engineer/supervisor for melt operations. Harvey Ulfers is a 29-yr veteran at John Deere Foundry with a bachelor's degree in business management and a master's degree in industrial technology, both from UNI. He is the supply management supervisor.