Late last year, Google received six hard disk drives from Seagate: brand new, save for their rare-earth magnets. The magnets had already spent one life in Google’s data centers and were headed for a second one.
The proof-of-concept experiment is one of five projects under a global initiative to examine the feasibility of recovering and reusing material from retired hard disk drives (HDDs). The program itself is a long-term effort organized by the International Electronics Manufacturing Initiative (iNEMI), with project team members from the U.S. Department of Energy's Critical Materials Institute. The results from all five projects will be summarized in a forthcoming report.
So far, early results are encouraging. “You can harvest [the magnets], you can do it cleanly, you can put it back together and the drives will work,” says Mark Schaffer, a consultant working with iNEMI.
The partnership grew from different goals that complemented each other. For many years, iNEMI had been looking into the sustainability of electronics manufacturing with a special focus on electronic waste. And, due to shocks in the supply of rare-earth elements in 2010, the U.S. Department of Energy started the Critical Materials Institute (CMI) to find ways to mitigate supply risks.
Finding new ways to extract critical materials like rare-earth elements from end-of-life electronics could serve as “one of our nation’s first lines of defense” against possible supply shocks, says Chris Haase, who leads the CMI based at Ames Laboratory in Iowa.
One “iconic product” was chosen in 2013 as a model for developing supply chains and business models to support a circular economy for electronics, says Carol Handwerker, a materials engineer at Purdue University who leads the metals recycling efforts at both iNEMI and CMI. That product was HDDs: ubiquitous components that are still relevant, particularly in data centers, but experience significant turnover.
By one estimate from Idaho National Laboratory, some 20 million hard drives are retired from major data centers every year just in the United States. Google and Microsoft, for instance, completely replace the HDDs in their data centers every two to three years, says Handwerker.
Those drives could be erased and reused, but most end up being shredded and sold for scrap aluminum or steel. In Tennessee, the leftover waste slag from metal recovery facilities, which contains rare earth elements including gold, platinum, and “all that good stuff,” is used as filler material for paving roads, says Tim McIntyre of Oak Ridge National Laboratory.
Now, a tasty carrot has appeared in the form of new stipulations in the Electronic Product Environmental Assessment Tool (EPEAT), a verification program for electronics produced using sustainable methods. Recently, the administrators of EPEAT began to award additional points for equipment with HDDs that contain at least 5 percent post-consumer content.
This is a big deal for companies bidding to supply IT equipment to U.S. federal agencies because those firms must earn EPEAT Gold ratings in order to win these contracts, says Handwerker. But for manufacturers wanting to purchase recycled rare earth materials to produce HDDs that meet those standards, there’s nowhere to go, says Preston Bryant of Illinois-based Momentum Technologies.
In two of the five projects under iNEMI's initiative, Momentum Technologies and Urban Mining Company are finding ways to scavenge magnet material from shredded drives. Using real-world shredded HDDs, Momentum researchers extract magnet material from the mix and turn it back into oxide powder—the starting material that magnet manufacturers use. Urban Mining Company, which is based in Austin, reprocesses recovered magnet powder into new magnet blocks.
Another strategy is to remove the valuable magnets whole. With Google’s six drives, the entire magnet assemblies were removed manually in a clean room by the Dutch electronics firm Teleplan, says Handwerker. The magnet assemblies were then passed on to Seagate, where they were placed into new drives. Google has since verified that the drives are functional.
However, for large-scale and high-throughput disassembly, automation is needed. Oak Ridge’s McIntyre has pioneered an automated way to remove magnets from hard drives. Most HDDs house their magnets in the lower left corner, says McIntyre, whose team examined more than 10,000 drives of some 250 different models and found only two exceptions.
The other thing to know for automated disassembly is where the drive fasteners are. McIntyre’s team created a database of fastener positions and programmed them into a machine. The machine scans a drive’s barcode, then removes the fasteners and cuts out the corner containing the magnet assembly. The corners are heated in an oven where the magnets fall out while getting demagnetized. Currently, a pilot facility at Oak Ridge National Lab can process up to 7,200 drives per day.
The magnets can then be reused in new HDDs—if they are not yet obsolete. Otherwise, they could also be repurposed for other applications, such as electric motors. A final option would be sending the whole magnets to companies like Momentum Technologies or Urban Mining Company for material recovery.
But by far the best scenario for value recovery in HDDs is reuse, says Handwerker. At Cascade Asset Management, a Wisconsin company that handles retired IT equipment, CEO Neil Peters-Michaud has shredders on-site but prefers not to use them. “If we can resell an item for reuse, it generates significantly more value than scrap commodities generate from shredding,” says Peters-Michaud, who participated in the life-cycle and economic analysis project under iNEMI's initiative. The problem with HDD reuse is companies are often concerned about data lingering after wiping. “Part of our discussion was, how do we convince organizations to trust electronics sanitation, rather than feel that they need to shred the drives?” he says.
Data wiping processes and building trust will be a main focus of the initiative’s next phase, says iNEMI’s Schaffer. Another goal is to extend recovery to other components, like circuit boards. That phase will also look at how to scale up successful proof-of-concept experiments and integrate them into real-world supply chains. “It takes a village,” says Handwerker.