Putting sand reclamation to the test at General Motors.

GM recently opened a prototype unit that could lay the groundwork for widespread sand reclamation at all its foundries.

As the once-easy route of landfilling or stockpiling spent foundry sand has hit roadblock after roadblock, metalcasters everywhere are examining options in reclamation,

Developing a strategy before the curtain falls, General Motors Powertrain Group unveiled a prototype unit at its Malleable Iron foundry in Saginaw, Michigan, this spring. Known as the Sand Reclamation Development Center (SRDC), it will validate the sand reclamation process in a production setting, and GM will use that knowledge to install units to handle reclamation for all its foundries.

Investing $5 million in the project, GM officials feel the 5-ton/hour ther-mal-mechanical reclamation unit marks a major step forward in addressing the issues of inadequate sand disposal space and increasing disposal costs.

"GM Powertrain saw this as an opportunity to install a prototype sand reclamation system," said Mike Williams, director of Manufacturing-Castings. "Due to the many different sand and binder systems used in GM casting plants, it was decided that a center must be established to prove technical and economic feasibility, ensure that product quality levels are not adversely affected and determine the requirements for corporatewide implementation."

Realizing that inevitable government mandates are expensive to implement, Larry Stahl, senior project engineer, said: "We want to be proactive and already be ahead in reclamation before they say we have to."

GM Casting

GM Powertrain's sand mold castings include engine blocks, cylinder heads, crankshafts, connecting rods, steering gear housings and steering knuckles. The North American casting operations at GM use 650,000 tons of sand/year, with the majority considered reusable. Through reclamation, GM officials believe they can ultimately reduce newsand use by more than 75%.

In addition to depleting sand resources from the western Michigan sand dunes, GM's on-site landfills are becoming full (Saginaw Malleable Iron already must transport its spent sand to the Saginaw Grey Iron landfill). It is difficult to obtain approval to build new land-fills, and landfill construction and closing costs are continuing to escalate. Liability is also a factor in operating landfills, since monitoring responsibilities may extend 50 years or beyond their closure.

Government regulations have caused the drastic increase of sand disposal costs, to the point where it is higher than the cost of new sand itself. On-site disposal costs are about $20/ton, off-site disposal costs range from $30-85, and new sand costs $10-20/ton.

Despite the economics, hooking GM up to widespread sand reclamation is no easy task. GM's larger North American casting plants, such as Saginaw Grey Iron, would require a sand reclamation system with the capacity to process 2530 tons/hour. While systems in Europe and Japan reclaim 12-15 tons/hour (due to high costs of sand and waste disposal), most U.S. sand reclamation systems reclaim at just 0.5-6 tons/hour.

Previous Efforts

GM has conducted many reclamation studies over the previous 35 years. Every GM foundry has, at one time, installed some type of sand reclaim system--all of which have been shut down and torn out. Basic requirements regarding reclamation were known, but concepts weren't proven.

Previous technologies to reclaim spent sand were prohibitively expensive and required tremendous amounts of energy, said William Shaw, director of Manufacturing Engineering.

"Multidisciplined teams weren't involved, maintenance systems were limited and often the systems were stuffed in a corner somewhere," added senior process engineer D.J. Couture. "There was no driving force behind the projects, since they could always landfill spent sand. Earlier, sand reclamation was expensive and caused quality problems."

So, conventional wisdom pointed to landfills as a cheaper and easier way to dispose of spent foundry sand. But as new sand and disposal costs skyrocketed and landfill space shrunk, reclamation began to make economic as well as environmental sense.

"Previously, our cost for new sand has been less than our cost to reclaim it," Shaw said. "Now, it's tilted the other way. For years, we've recognized it doesn't make sense to landfill sand. Now, we have the technology that we hope will allow us to reuse sand as much as possible, and limit our purchases of new sand."

Current Project History

The newfound driving force toward establishing the SRDC came in the late 1980s, when GM implemented a vacuum-assisted casting (VAC) line. Because the VAC process relies on large amounts of sand (up to 32 tons/hr), sand reclamation would be the only way the VAC process could operate competitively.

Experience showed that mechanical systems adequately reclaim green sand for green sand molding, and thermal or mechanical systems work for reclaiming chemically bonded sand to chemically bonded molding, Couture noted. But to reclaim green sand for chemically bonded molding, as GM requires, a combination of thermal-mechanical reclamation was needed.

Therefore, GM officials approved the SRDC (completely operational facility serving GM Saginaw Malleable Iron Plant) in 1992. The center is studying reclamation while determining requirements for corporatewide implementation. It was approved as an environmental project, rather than on a cost-savings basis, even though GM officials say they will eventually realize cost savings through this effort.

Development teams were formed to handle system design, installation and support. These teams include skilled United Auto Worker (UAW) employees and salaried employees representing areas such as engineering, materials management, environmental engineering, personnel and accounting.

"The system was designed to allow the most sophisticated testing," Stahl said. "We needed key process parameters and data. We had to cover as many details as possible and still have a system for real production. We didn't want to put it together and then throw our work away later."

Talking about the team's role in designing the system, Ron Cafferty, staff engineer, said years ago the relationship with the union nowhere resembled what it is like today: "Now it's a team approach. We bring hourly guys in at every step and seek their involvement. The level of hourly involvement has really ramped up. On a recent facility tour, in which D.J. Couture was away, a question was asked and one of the hourly guys said, 'Let me talk.' When he was asked questions, he answered them all."

Selection Process

After team members met with 12 then mal equipment vendors from around the world, they narrowed their selections down to four. The team used a weighted rating system with an initial survey of 17 criteria known as the "Concept Development Process," which provides an organized approach to concept evaluation and helps reduce personal bias in selection.

"It is a statement of work from someone's idea to the product," Couture said. "A logic network is followed with six to seven reviews with company officials, in which they can stop and revise the design or continue on. It sets up hurdles to make sure everything is addressed."

In the final equipment selection process, the team surveyed more than 200 criteria. Topics included management, maintainability, cost and test results. "They used a systematic approach in assessing all vendors," Cafferty said. "Suppliers knew how they would be judged up front, and they couldn't claim preferential treatment."

For instance, in selecting the appropriate calciner (the heart of the system), they tested a rotary kiln, open flame/batch system and two fluid-bed systems. Extensive testing at each supplier showed no significant difference in mold tensile strength, but fluid-bed systems had the lowest emission levels. In regard to initial casting results using the VAC process, again it was found there weren't differences between new and reclaimed sand. Therefore, the fluid-bed design was selected.

The SRDC

Located next to the Saginaw Malleable Iron foundry, the SRDC is an abandoned sand preparation building (80 x 90 x 50 ft) that was gutted out. Among the key features of the system are the fluid-bed calciner by ALB. Klein Co., a crusher from Didion Manufacturing Co., a magnetic separator from Eriez Magnetics, a gas train system from North American Manufacturing Co., and a refurbished dust collector, transferred from GM's Corvette plant.

Basically, the 5-ton/hour system works as follows:

* sand is collected;

* rotary dram lump crusher reduces size of material to minus 1/8 in.;

* a magnetic separator removes metallic particles from the sand;

* the fluidized bed furnace calciner removes binders through natural gas combustion at 1400F for 1 hr;

* after cooling, fines are removed that will be disposed, sold or beneficially reused.

* reclaimed sand is transported to final processed sand silos.

Other equipment that will soon be installed at the SRDC for green sand reclamation include:

* dryer--required for sand with over 2% moisture;

* classification system--equipment to split the reclaimed sand into multiple size ranges (coarse, medium and fine) to allow for custom blending of the sand to match the specifications of the multiple sands used;

* production-sized version of selected clay removal system--critical to the success of green sand reclamation.

Product Quality

Of the lab and casting testing conducted thus far, there has been no major difference in tensile strength of the sand or metal penetration of the reclaimed sand vs. new sand. Production rate demonstrations are under way.

Besides the environmental and economic benefits, reclamation also gives the foundry the sand it needs. "When you buy sand, you get some undesirable elements," Stahl said. "Now, we get the sand the foundry really wants--consistency is the name of the game."

Couture added, "We retain about 95% of the sand each time through the system, and it's a better quality sand. It's the survival of the fittest."

Implementation

Development work by the team includes these goals:

* validate thermal reclamation with chemically bonded sand;

* evaluate and test clay removal methods;

* validate the total sand reclamation process for green sand to the core room (test each binder system and sand from each plant);

* determine requirements for production systems;

* transfer technology to other GM foundries.

The team noted that the keys to GM's reclamation success will be: ability to segregate waste streams; binder system compatibility; minimizing additives to sand (iron oxides, inorganics); clay removal effectiveness for green sand; and realizing that GM is becoming its own sand supplier.

GM hopes to conduct the initial scrubber evaluation this summer, and then install it this fall. They intend to validate green sand by the summer of 1995. If the process is validated so that it improves or maintains necessary levels of quality, company officials will begin planning for full-scale production reclamation systems in 1996.

"We will develop the process, validate it and look at the paybacks," Couture said. "When it makes sense, GM will go ahead with full reclamation."