Rust: The Longest War

by Jonathan Waldman

  But that perfect end remains susceptible to corrosion. That’s why the snowbird move is so detrimental. The only thing touching the inside of the end ought to be harmless carbon dioxide (or nitrogen). Leave a can sideways for six months, and it’ll leak. “I’ve often wondered,” Scheuerman told me, “why they don’t write ‘Store upright’ on the cans.”

  It took two years of research before I met anyone who had experienced an exploding can. As it happened, the man who told me of that unlikely event wasn’t a snowbird but Jamil Baghdachi, of Ypsilanti, Michigan. In the summer of 2006, Baghdachi was on his way from Louisville, Kentucky, to the airport. It was rush hour, and just as he drove his Volvo under an overpass, a soda can on the passenger seat exploded, spraying all over his face. To Baghdachi, it was not a disaster. He is the director of the Coatings Research Institute at Eastern Michigan University. He has consulted on can coatings and has dozens of coatings patents. As he explained, “It was a lack of a complete coating.”

  Two hundred years ago, it took an hour to cut, assemble, fill, and solder shut the first cans. And then, because there was no other way to open them, people attacked their cans with knives, bayonets, hammers, and chisels. They smashed them open with rocks or fired rifles at them. For fifty years, even though their contents had often spoiled, cans seemed like marvelous containers. Can openers were invented, and cans seemed even better. By the turn of the twentieth century, with can-making machines spitting out one hundred cans a minute and pasteurizing techniques down pat, the can seemed incredible as ever. A generation later, so did Ball jars, but you could get beer in a can. Surely this metal container could improve no further.

  A generation later, this container was made of aluminum. Not long after, manufacturers learned to crank out a thousand of them a minute. They learned to make them thinner and lighter, openable not with a hammer or a church key but with a riveted tab and scored panel. They learned to crank out two thousand of them a minute. They learned to line them with such durable plastics that beverages of nearly unimaginable corrosiveness could be put inside. They painted them in thermochromic inks, lined them with blue coatings, devised resealable screw tops. Having made trillions of these things, Ball now makes cans accurate to 50 millionths of an inch, and aims for errors to be so rare that they are six standard deviations from the mean. The only thing Ball can’t make is clear aluminum, so that customers might see the water they intend to put inside. Cans have improved so much that engineers at Ball—self-declared perfectionists—debate whether or not they can improve any further.

  On the second day of Can School, Mary Chopyak, who has spent twenty-one years as a materials engineer at Ball, pointed to the asymptotically shaped curve of can weights over the last fifty years, and announced that we’re at the tail end of improvements. In the last twenty-five years, cans have only gotten one-hundredth of a pound lighter. She said that there are only minuscule improvements left to be made in the can. Sandy Deweese, an engineer who works on can ends, said, with a slight drawl, that nothing else can change. He said that can engineering has reached its limit. Yet Dave Wrenshall, a technical advisor at Ball—tough, mustachioed, somehow reminiscent of Robert De Niro—said it’s hogwash; that we’re not near the end of optimization. He kept his feet squarely planted, as if there were magnets in his shoes. “People have said that for the last twenty years,” he said. He didn’t mention foreign objects or explosions, or epoxy internal coatings.

  When I first heard about Can School, I called a public relations employee at Ball named John Saalwachter. Saying that he was excited for me to attend, Saalwachter put me in touch with his boss, Ball’s director of corporate relations, Scott McCarty. This was a couple of months ahead of time. Then I did some preparation and asked a few too many questions. Emailing Ball’s general counsel, I asked about the various liability concerns relating to foreign objects and explosions. This probably raised some flags. I also inquired about the internal coatings, and this certainly raised more. Two weeks before Can School, McCarty called me. In no uncertain terms, he told me I couldn’t come. He said that Can School wasn’t for journalists; that no journalist had ever attended. That it wouldn’t be fair to those in the industry.

  It wouldn’t be fair to the industry, I came to conclude, because the internal coatings used to prevent aluminum cans from rusting from the inside out and adulterating their products are as secret and controversial as fracking fluid. The formulas used by the major coatings manufacturers—PPG Industries, Valspar, and Akzo Nobel—are proprietary; the details vague. I called each of the major manufacturers; none returned my calls. In legal documents, details are carefully marked confidential and redacted. In patents, formulas are left vague—citing, for example, a component added in a range anywhere from 0.1 percent to 10 percent. Even the US Food and Drug Administration (FDA) must, as one chemist there put it, “pry stuff out of” the manufacturers. Often the formula is unpatented and therefore preserved as a trade secret.

  This much is clear: the epoxies must be affordable, sprayable, curable, strong, flexible, and sticky—such that not much beyond sulfuric acid or the potent solvent methyl ethyl ketone (MEK) will remove them. Creating such a material takes a cross-linking resin, curing catalysts, and some additives to give it color or clarity, lubrication, antioxidative properties, flow, stability, plasticity, and a smooth surface. The resin is usually epoxy, but it may also be vinyl, acrylic, polyester, or oleoresin, and could even be styrene, polyethylene, or polypropylene, or a natural drying oil derived from beechnut, linseed, or soybeans. The mixture also requires either a solvent, so that the epoxy can cure when baked, or a photo-initiator, so that the epoxy can cure when exposed briefly to ultraviolet (UV) light. The cross-linking agent of choice for the most tenacious epoxy coating is bisphenol-A, or BPA. BPA is the primary ingredient in such coatings because it makes the plastic plastic.

  In addition to thickness, which is determined by the corrosivity of the product, beverage can coatings engineers pay attention to how the can will be treated. Will it be pasteurized? Stored in a hot place? Shipped on a vibrating train through someplace hotter or colder or more or less humid? They tweak the application or select a different coating as required. For beer, engineers might use a coating, patented by Valspar, that contains cyclodextrin, a donut-shaped carbohydrate that, via molecular inclusion, traps bad-tasting molecules. Food can coating engineers have greater concerns. Coatings for tomatoes must be stain resistant, those for fish must resist sulfur, and those for fruits and pickles must resist acids. There’s one coating for tomatoes, one for beans, one for potatoes, and another for corn, peas, fish, and shrimp. Chocolate, especially sensitive to adulteration, requires its own coating. Those for meats must contain a lubricious wax, called a meat release agent, so that the meat slides right out. Fruits and vegetables including beets, currants, and plums, which contain the red pigment anthocyanin, are some of the most corrosive. The top of the list belongs to rhubarb. It, alone among foods, requires three layers of lacquer, and even with that much protection, rhubarb still boasts a shorter shelf-life than its peers. All told, there are over fifteen thousand coatings, and though most of them serve inside food cans, many of them perform their work inside beverage containers. Beverage cans in the United States demand about twenty million gallons of epoxy coatings every year, for about one hundred billion cans. According to coatings specialists, roughly 80 percent of that epoxy is BPA. A tiny bit of it, then, ends up in us—reason, perhaps, for the secrecy.

  Biologically speaking, hormones are rare, and potent. The system that produces, stores, and secretes them—the endocrine system—controls hair growth, reproduction, cognitive performance, injury response, excretion, sensory perception, cell division, and metabolic rate. Endocrine organs—including the thyroid, pituitary, and adrenal glands—produce particular molecules that fit into particular receptors on cells, unleashing a chain of biochemical events. Hormonal changes in infinitesimal quantities cause dramatic changes, including diabetes and hermaphrodites. End
ocrine disruptors, including molecules that mimic the hormone estrogen—called estrogenic chemicals, or xenoestrogens—get jammed in the cells so that the real molecules can’t get in there and do what they should. Others fit perfectly, triggering events the body didn’t intend to initiate.

  Clues of synthetic chemicals having such effects, detailed in Rachel Carson’s 1962 book Silent Spring, had been piling up since the 1950s. Beginning in the late seventies, wildlife biologists around the Great Lakes, primarily studying fish and birds, found that compounds were altering the cells, bodies and behavior of the animals in novel ways. They found masculinized female fish, feminized male fish, intersex fish, and birds that refused to raise their young. In a landmark 1993 study, biologists Theo Colborn and Frederick vom Saal described the range of this “endocrine disruption,” a term coined only two years earlier.

  BPA’s role as an endocrine disruptor wasn’t recognized until 1998. That’s when Pat Hunt, a geneticist at Case Western Reserve University, in Cleveland, noticed something strange about an experiment she was running on some mice. Forty percent of her control mice—the ostensibly normal, healthy ones—were producing abnormal eggs. “We checked everything,” she told the writer Florence Williams. After weeks of ruling out potential culprits, including the air in the lab, Hunt noticed smears and scratches on the animals’ plastic cages. It turned out someone had used an acidic floor cleaner, rather than a mild detergent, to clean the cages, and it was degrading them. Something was leaching into the mice’s feeding tubes and interfering with the mice.

  The something was BPA, an artificial estrogen first studied in the 1930s to prevent miscarriages. It didn’t work for that—it has quite the opposite effect—but the double-hexagon-shaped molecule was soon put to work making polycarbonate plastics shatterproof, which made them the future, and abundant. It took a half century, but scientists—finally studying chronic toxins rather than acute toxins—caught up with plastic. A 2011 paper titled “Most Plastic Products Release Estrogenic Chemicals,” published by the National Institutes of Health’s National Institute of Environmental Health Sciences, sums up the current understanding. In it, researchers describe detecting estrogenic activity in over five hundred commercially available plastics, including many advertised as BPA free. They report that manufacturing processes—such as pasteurization—convert nonestrogenic chemicals into estrogenic chemicals, and they note that sunlight, microwave radiation, and machine dish washing accelerate the leaching of estrogenic chemicals. They also mention that the most effective way to release estrogenic chemicals is by using a mixture of polar and nonpolar solvents (in their case, salt water and ethanol), which pretty much describes Bob’s Energy Drink.

  According to Hunt, who doses mice with amounts of BPA proportional to their body weight, she sees abnormal behavior or cells in the mother, her offspring, and her offspring’s offspring. One dose damages three generations. This is because BPA shows up in cells in the breast—the body organ most sensitive to the effects of estrogen—which means that its effects get passed to the next generation at the most fragile, sensitive time. Timing turns out to be a major factor. In fact, it turns out that exposure to BPA may be comparable to exposure to DDT, the carcinogenic insecticide about which Carson wrote. Girls exposed to DDT before puberty, researchers found, have five times the cancer rate of girls exposed to the same amount afterward.

  Hunt, Colborn, vom Saal, and many other researchers around the world have found that BPA can cause early puberty, obesity, and miscarriages, lower sperm counts, and increase rates of cancers of the breast, prostate, ovaries, and testicles. All these troubles were observed in rodents, many of which were exposed while in their mothers’ wombs. In other rat experiments, prenatal exposure to low doses of BPA caused lesions in mammary glands. Other studies have confirmed that BPA activates the estrogen receptors on breast cells, and can cause cancer cells to replicate in a dish. BPA has been shown to cause normal breast cells to act like cancer cells. It’s also been shown to be just as powerful as diethylstilbestrol, or DES, a very strong synthetic estrogen discouraged since 1971 in pregnant women because of its association with rare, awful reproductive cancers.

  Hence, perhaps, the retraction of my invitation.

  After McCarty told me to steer clear, I was devastated. I spent a morning searching for people who’d been to Can School (it’s amazing what people put on their résumés) and the rest of the day trying to get in touch with them. I emailed a friend at a local brewery and asked if any of her colleagues were going. I got nowhere. The day after McCarty’s call, I received an email from Clif Reichard. The subject was Ball Beverage Can School. I figured it was something from a lawyer. The first sentence of the email said, “We have you registered to attend our Beverage Can School.” I read it a few times before continuing. It went on, telling me where and when to show up. It said that lunch was included, that dress was business casual, and that open-toed shoes were not allowed for the tour of the plant. It had Reichard’s cell phone number. And it thanked me, ahead of time, for coming. I didn’t know what had happened. Maybe the higher-ups at Ball had changed their minds. Maybe they goofed up. I didn’t care. I was in. Twelve days later, I drove down to Broomfield.

  Ball’s Broomfield headquarters, halfway between Denver and Boulder, is situated on an awesome piece of property, offering a full hundred-mile panorama of the Front Range of the Rocky Mountains. From north to south, Longs Peak, Eldorado Canyon, Mount Evans, and Pikes Peak are visible. Driving in, I was nervous. I noticed two video cameras on the south side of the driveway, just after a sign that said PRIVATE DRIVE. I wondered: Would lawyers pounce on me? Would security? Might I get escorted out? Or arrested?

  From the parking lot, I walked up a path, under some trees, and through two sets of dark glass doors. The lobby within was a sunny atrium, with a dozen leather chairs in the middle. On the right, there was a flat-screen TV with Ball’s stock price ticking across the bottom. It was $36.15. On the left, there was a security desk, and past it, an enlarged sepia print of the five Ball brothers looking all business, their mustaches lifesize and resplendent. I headed for the security desk.

  Sweating, I tried to play it cool. I said I was there to attend Can School. The attendant asked for my name and began scrolling through a stack of ID cards. What was my name again? He couldn’t find my card. He asked for a driver’s license and had me enter my name on a sign-in sheet. Then he handed me a temporary badge, which said Escort Required. Being badgeless seemed advantageous. I could be anybody, or nobody. Still, I stayed low, not flaunting my presence. The attendant pointed me down the hall, to a conference room. He warned me not to deviate without an escort.

  I walked in, and sat in the second row, near the right, by myself. Nervous, I sat there as if departure was imminent. I kept trying not to look back over my shoulder. I tried to listen to the hushed small talk around me. I was more paranoid than I am proud to admit. Ten minutes later, a security guard brought me a real name tag, and a nameplate for my seat. Below my name, it said Scripps—the name of the journalism fellowship I was on. It seemed vague enough, almost like a brand of soda. Five minutes later, Clif Reichard came over, said sorry, and said he’d give ’em hell for taking me off the registration list. I said, really, it’s no big deal, I live just up the road. Really, it’s no problem.

  During the initial presentations, I didn’t ask many questions, because I wanted to stay under the radar. I later gave in and chatted up two employees from the Can Manufacturing Institute. Megan Daum, CMI’s sustainability director, was apprehensive. Her body language made it clear that she wanted out. Joseph Pouliot, CMI’s vice president of public affairs, was chatty, and smart. When I pressed him about CMI’s unremarkable motto that cans are “infinitely recyclable,” he admitted that glass, too, fit the bill.

  By the end of the day, I was worried again that lawyers awaited. Paranoia was getting to me. Would they demand my notes? My recorder? The thumb drive and documents they’d provided? Not wanting to write that
in my notes, I wrote in Spanish, scrawlier than normal: “puso memoria en pantalones, y uso español en mi papel,” which was my tired way of noting that I’d put the thumb drive down my pants and used Spanish to conceal that fact.

  A good coating is hard to find, and as much effort goes into creating and applying one as goes into any other aspect of manufacturing cans. This was not taught at Can School.

  Making an epoxy coating starts with a petroleum refiner like ExxonMobil, which produces huge quantities of benzene. Dow Chemical Company or Momentive Specialty Chemicals converts the benzene to bisphenol-A, and combines it, about 4:1, with epichlorohydrin. Dow calls this stuff D.E.R. 331, or Dow Epoxy Resin 331, and Momentive calls it Epon 829. These blends, among dozens, are suitable foundations for coatings on everything from cans to cars and bridges. Next, a chemical company such as Cytec Industries or Sartomer or Rahn buys the epoxy resin, adds 5 percent acrylate to make an acrylated epoxy called something like Genomer 2255, and sells it to one of the big coatings companies. To render the coating just right for Bob’s Energy Drink, the major coatings companies then add small amounts of pigments, surfactants, adhesion promoters, corrosion inhibitors, light stabilizers, toners, extenders, thixotropic agents, dispersing agents, wetting agents, dyes, and catalysts—which are often made by yet another chemical company. With their own IC spray machines, the coatings companies line cans, and make sure that the coating doesn’t feather when the tab is lifted, craze when dented, or blush in the presence of certain solutions. The finished product costs about $25 per gallon.