The Powerhouse: Inside the Invention of a Battery to Save the World

by Steve LeVine

  At the ARPA-E Summit, Kumar and Kapadia had led a small news conference that included a slide with the first Crane results. Any journalist, if sufficiently acquainted with cycle testing, could study the slide and write the dismal truth about the Envia cathode. But none appeared to be so informed and, absent Kumar or Kapadia calling attention to the plunge in energy density, it went unreported. Kumar’s sin of omission crossed the line into explicit deception the same day, when Envia distributed a news release on the 400-watt-hour-per-kilogram breakthrough. This announcement excluded the Crane report entirely and skipped its description of the plummeting energy density. Instead, it published charts that were not part of the Crane evaluation at all. These charts described the anode, which, the illustrations said, continued to work at 91 percent of capacity after three hundred cycles, a very impressive number. The chart was deceptive not only because it displayed data on which Crane did not report, but because it suggested that the cell as a whole was relatively stable.

  The trick was the chart’s subtle substitution of the metric capacity, rather than energy density, what Crane measured. Capacity is a measure of how many electrons flow in a single discharge of the battery cell. It can be maintained for numerous cycles even as the cell’s energy density is plunging. Capacity was precisely what had deceived Thackeray and Johnson when they were touting their invention of NMC 2.0. The energy of each individual electron—in essence what the voltage measures—was lessening each cycle. This was another way of describing voltage fade. At the three hundredth cycle, the Envia cell’s energy was at 60 percent of its original level and dropping, nowhere close to the 91 percent of capacity that was registered.1

  For twenty months—from the time of ARPA-E until the fallout with GM resulted in public release of the Crane report—Envia released detailed charts depicting only the capacity. It allowed the impression to persist that it was achieving energy density of 400 watt-hours per kilogram for hundreds of cycles. Again and again, Kumar created the appearance that the cell was fine when the energy density was going haywire. Only when the news of the fallout with GM went on the Internet did he change the company Web site, pull down the charts on capacity, and post the second Crane report. He never released the complete first Crane report—the one on which the ARPA-E announcement was actually based.

  Yet that was not the total picture. There was more to Kumar, and when you looked at everything, his behavior did not resemble a pure confidence game. His pursuit of Jason Croy did not square with a picture of a con man. His frustration with the national labs also lacked the appearance of a scam. Kumar genuinely thought he could parlay the Argonne material into a blockbuster new battery and enable the electric-car industry. History is filled with single-minded entrepreneurs with outsized notions of their prospects. He also was chasing an outsized payday. Along the way, Kumar understood that he was in over his head. But he had made too many public and private claims. If he just held on, the payday could come. In the end, as the wreck was approaching, he hoped that somehow everything would work out. Even if he could not reproduce the 400-watt-hour-per-kilogram super-battery, nor the lesser 350-watt-hour-per-kilogram version, perhaps GM would accept something less.

  As for Kapadia, he claimed he had known nothing until the end, but he clearly had. He knew of the Crane results, which he himself described to journalists. He was prepared to live with the deceptive anode: at least twice in the GM crisis, he offered his resignation but never walked out the door. His team, too, wanted to stay on despite the faulty performance, the swapped anode, and the lost GM contract, but were not permitted to.

  The board, made up almost entirely of investors and executives, had personal incentives to stay quiet and hope that Kumar somehow still managed to win so that they could cash out. Which is what they did.

  As Gallagher indirectly suggested, GM was party to this excessive enthusiasm. Disregarding its engineers, its VC and technology sides thought their long, painstaking vetting of Envia’s coin cells was sufficient precaution against product risk. But it wasn’t. The carmaker, presumably in possession of the conclusions from the Crane report, decided that in just twenty months Kumar could solve electrochemical problems that were vexing some of the best labs in the world. That was a lot of hope to hang on Envia.

  Arun Majumdar, the showy ARPA-E director who had acclaimed the start-up to the industry and the world, left the agency and joined Google. In early 2013, he also became a member of the Envia board. After it all fell apart, Majumdar said the attempt to put Envia’s work into a commercial vehicle was premature. “Envia is not ready for prime time in terms of product. That’s been the realization,” he said.

  As for why Lauckner and the rest of GM elected to stay mum and maintain the appearance of a relationship with Envia after the fallout, that was understandable as well. There was no profit in going public with a fiasco. The news could discredit the Volt and tarnish GM’s reputation for developing new technology. Wall Street could pummel the stock. Politically the fallout would also be uncomfortable, as the Republican Party would rub both GM’s and the Obama administration’s nose in what it would call a waste of millions of taxpayer dollars on useless stabs into clean energy.

  Wan Gang was still out there, as were Japan and South Korea. The big stakes remained. So even in a circumstance of misrepresentation, exaggeration, sleight of hand, and general slipperiness, for GM there was a case to be made for sending Kumar back to the lab. He was a scientist. The remote prospect remained that some time in the future, by a miracle, Kumar actually would succeed.

  • • •

  In the weeks after, Atul Kapadia and his two lieutenants asked the Envia board to issue a short press release explaining their departure as a change of corporate strategy and thanking them for their work. If Envia could not see fit to do so, they attached a fifty-two-page civil suit they intended to file in the Alameda County Courthouse against Envia and Kumar personally. It alleged fraud, wrongful termination, and retaliation for their trying to right the false representations about the anode. When the company was slow to respond, the men filed the suit, divulging many of the previous six years of corporate secrets. It was yet another Envia blunder. The mission of keeping the final, sorry year under wraps was blown.

  Kapadia’s lawsuit borrowed extensively from the legal complaint filed the previous year by NanoeXa, the company where Kumar and his cofounder Mike Sinkula had worked prior to Envia. When examined together, the suits alleged that Envia, while claiming to be marketing proprietary technology, relied almost entirely on IP that it either stole or appropriated without attribution from NanoeXa and Shin-Etsu. In terms of the cathode, the suits, relying on the work of a private forensic investigator, alleged that Kumar secretly downloaded ninety-nine files from the computers at NanoeXa, many of them in a flurry over his last days and hours at the company. Without these files, the suits asserted, there would have been no Envia as it came to be known. Envia flatly denied the NanoeXa assertions. About this time, Kapadia’s wife suffered a relapse of cancer, and he and the other two former Envia executives dropped the lawsuit.2

  46

  Back to the Race

  After Argonne won the Hub, Chamberlain barely had time to think about celebrating. The Department of Energy insisted on an entirely revised Hub budget—immediately. He thought back over the three years since he first floated the idea for a research effort on the scale of Bell Labs, recalling the history as though telling it for himself as much as anyone else—En-Caesar, the Battery Sematech, and briefing the Obama campaign. He reached back to Cabot Industries, his early experiences in the private sector, and the day he left the start-up world thinking he would coast into Argonne’s intellectual property unit for a few relaxing years.

  In the ensuing months, Chamberlain would learn that, just as American plans had suffered a setback, Wan Gang had tempered his own aspirations and shifted direction. China’s motorists continued to adopt electrics even more slowly than did Americans—so
me 80 percent of the forty thousand electric vehicles on China’s roads in the coming year would be buses and taxis. Wan said he would eliminate a $10,000-a-car subsidy entirely by 2020. The government would instead offer grants to company research labs.

  Wu Feng, a senior Chinese official who visited Argonne in late 2013, said that when Wan took up the battery race, he was an electric car expert. China’s battery guys had warned him that energy storage was “difficult.” But he and the other car specialists weren’t absorbing the message. “They imagined that it is easy.” Now Wan realized that “it will be a long time to reach the goals.” China was still in the race. But Wan had learned that the country would not win quickly.

  Some members of the American team felt the same. Peter Faguy, the Department of Energy official who had ordered the crash effort to solve voltage fade, was shutting down the team. Would fade ever be figured out and NMC 2.0 fully enabled? Perhaps, Faguy said. He was not opposed to anyone trying again. But voltage fade had eradicated his own hopes that you could throw money and desire at a problem. “The romantic notion of the line researcher solving it with the light bulb going off is a nonstarter,” Faguy said. “These kind of problems are intractable.”

  • • •

  Some even said the race was misconceived from the start. Economic and technological hope and coincidence led a lot of nations to chase an illusory prize that then evaporated in their hands. When it was time for the better battery and electric cars, they would arrive and spread quickly, with dividends crossing borders and no single national winner. But that time was not yet with us.

  Don Hillebrand disagreed. Extremely capable scientists backed by patriotic governments with concrete and reasonable objectives had made a sincere dash for a better technological path, one that held almost magical powers to resolve some of the era’s most intractable economic, political, and environmental problems. When hopes and the stakes for winning are so great, however, you can bring out not only those with potential answers—visionaries—but also “charismatic thieves, swindlers, who are tricking people,” Hillebrand said. And, in the case of the battery, when you combined those tendencies with the perseverance of the internal combustion engine, you got a race that “ended in the middle.” But Hillebrand predicted that the Hub would invent a blockbuster new battery chemistry. It was a matter of survival. The Hub scientists “know if they don’t, they’ll be in trouble.” Specifically, Chamberlain’s “career is very much on the line and he knows it.”

  The War Room was closed. Chamberlain moved into a modern, freshly carpeted suite of offices in Building 200. His new title was Deputy Director of Development and Demonstration for the Hub. That meant that he was responsible for delivering the prototypes promised under the five-five-five criteria. He said he was “highly confident” that he would do so and thus create a new paradigm for American manufacturing. It would be Bell Labs 2.0. He said, “I’m hoping in five or ten years to be touring the country saying, ‘This is how it can be done.’”

  He watched electrics quietly moving ahead. The Volt for sure was a pioneering vehicle, but Elon Musk had pushed further—he had made electrics indisputably cool. With Tesla, Musk himself was now the most celebrated technologist in Silicon Valley.

  Toward the end of 2014, a mini-rivalry erupted: Musk hurtled into a contest with GM to produce the two-hundred-mile electric. He did not say he was in competition with GM—in his eyes, that would be demeaning. But after vowing for years to produce a mass-market electric by the end of the decade, he now said he would do so in 2017. He said the car, to be called the Model III, would cost about $35,000. And that fact—the price—put Musk squarely on GM’s turf.

  Jon Lauckner, still smarting over the Envia debacle, did not say when GM would release its own two-hundred-mile electric, but it would not easily relinquish its market. GM did not desire a direct clash with Musk, given his rare mastery of product style and marketing—his pizzazz—but it had one. If he intended to be out with his Model III in 2017, GM would have to have its rival model on sale that year or earlier.

  The stakes were clear. The top electrics—the Volt, the Leaf, and Musk’s Model S—were selling at a pace of 2,000 to 3,000 vehicles a month each, but motorists were buying about 40,000 of the BMW 3-Series, the entry-level gasoline-driven luxury car that Musk identified as his genuine competition. They would be somewhat over the price of the average vehicle, but at that rate of sales, the competing GMs and Teslas would tip electrics into the broad consumer market. They would no longer be niche vehicles. At once, Obama’s aim for 1 million electrics on the road would be realized. And that is what Musk said he would do—he alone would sell 500,000 electrics a year by 2020.

  LG Chemical, GM’s lithium-ion battery supplier, contributed to the drama. Prompted by no one obvious, a senior executive blurted out at an earnings presentation that the company would manufacture a two-hundred-mile battery in 2016.1 He did not say for whom LG was making the battery, but the disclosure seemed to further telegraph GM’s forthcoming battle with Musk.

  The pieces were in place. Tesla and GM would achieve their range by re-engineering current-generation battery technology and lightening up their respective vehicles. The Hub, by starting over from scratch, stood a fighting chance of taking batteries the rest of the way.

  Chamberlain said the Hub meant peering deep into the physics, with no assumptions as to where the answer lay. Armed with an atomic road map of the chemistry, the United States could really win.

  Acknowledgments

  When I began to consider a book on batteries, the reception from friends and advisers was all but unanimous: don’t do it. It would be a tedious bore. The exceptions were a few hands at Argonne National Laboratory, who seemed to understand the fascination instantly. They are Jeff Chamberlain, Angela Hardin, and Don Hillebrand. I want to single out Chamberlain, without whose support this book would not have happened. The decision to welcome a stranger with a tape recorder into a secure setting with colossally high stakes was not trivial. I would ultimately spend two years in the Battery Department, sometimes hanging around for a few days and once for two months. Chamberlain let go, gave me the run of the department, and persuaded the rest of the lab to relax. They allowed me to be the fly on the wall that I had hoped. I am grateful to Eric Isaacs, Argonne’s director, whose permission was required, and David Sandalow, who made a timely call to Isaacs on my behalf that finally set everything in motion. I am also grateful to Sujeet Kumar and Atul Kapadia, who welcomed me to observe Envia’s quest to be a successful start-up battery company.

  Thanks to New America, which supported this book in an astonishing three-year-long fellowship. New America is a unique institution whose literary and intellectual sensibilities constantly result in outstanding work. Specifically, thanks to Steve Coll, Andres Martinez, Rachel White, and Anne-Marie Slaughter. Superlative thanks to Kevin Delaney, my editor at Quartz, a gentle chief who runs the best publication on the Internet and understood my fixation. In the first fourteen months of Quartz’s existence, Kevin accorded me roughly ten months of either part- or full-time book leave. Certainly no other boss would have been so generous. Thanks also to David Bradley, who demonstrates that it is simultaneously possible to be a media colossus and a nice man. I cannot say enough about Jim Levine, the most trustworthy agent and friend in publishing. Jim read and commented on the manuscript three times. Thanks to Rick Kot for embracing the book and to Melanie Tortoroli for her artful and supple advice and editing and ultimately her enthusiasm—she is a terrific friend. Additional thanks to Viking’s Ben Petrone and Sarah Janet for their superb work. Susan Glasser commissioned the article for Foreign Policy that led to this book and Charlie Homans edited it. Alyson Wright, my research assistant and transcriber, diligently listened to, typed out, and checked hundreds of hours of interviews and offered wise counsel on the content. Noel Greenwood gets an exceptional expression of thanks. Noel edited my first two books and committed to do so this time as well.
Throughout, he was in treatment for cancer. Finally, Noel decided enough was enough. I saw him a week before he died and we settled on a road map for The Powerhouse, one that I stuck to pretty closely. Thanks, Noel.

  I am solely responsible for any errors but I am grateful to Billy Woodford of MIT, who read the full manuscript three times and returned with pages of important scientific and technical corrections, tweaks, and suggestions. Christian Caryl, Konstantin Kakaes, Chris Leonard, Sharon Moshavi, and Chris White read the manuscript and saved me from numerous errors in addition to offering important stylistic suggestions.

  A special thanks to those who had the patience to explain (often repeatedly) the science and history of batteries, technology, big geopolitics, and Argonne (with apologies in advance for the names I will no doubt omit for no reason apart from personal shortcomings): Khalil Amine, Jason Croy, Kevin Gallagher, Don Hillebrand, Mike Thackeray, Lynn Trahey, and Brad Ullrick. Also Daniel Abraham, Ralph Brodd, Emilio Bunel, Tony Burrell, Yet-Ming Chiang, George Crabtree, Jeff Dahn, Sun-Ho Kang, Chris Johnson, Peter Littlewood, Paul Nelson, Mark Peters, Venkat Srinivasan, and Jack Vaughey. In addition, Ali Abouimrane, Elizabeth Austin, Mali Balasubramanian, Ilias Belharouak, Martin Bettge, Benjamin Blaiszik, Terry Bray, Rita Brzowski, Zonghai Chen, Holly Coghill, Larry Curtiss, Bill David, Dennis Dees, Carolyn Edmonson, Dan Flores, Sharon Giblin, Gary Henriksen, Matt Howard, Donghan Kim, Greg Krumdick, Jun Lu, Nenad Markovic, Vilas Pol, Yang Ren, Dave Schroeder, Mike Slater, and Cynthia Sullivan.