CLEAN TECH: SPECIAL REPORT
Legendary venture capitalist Vinod Khosla backed a startup called KiOR as part of his ambitious push for green energy. Now KiOR is bankrupt, he and company executives are being sued for fraud, and Khosla’s big biofuel bet is looking increasingly questionable.
Drive a couple of hours northeast of Jackson, Miss.—maneuvering past semitrailers piled with timber on highways lined with dense pine forests—and you’ll reach a sprawling structure of metal pipes and towers.
It’s a factory in the sleepy city of Columbus, Miss., that was briefly able to turn wood chips into a biofuel that could power vehicles more cleanly than oil, using a first-of-its-kind technology that wowed engineers, politicians, and investors.
The plant’s supporters once envisioned it as the embodiment of a clean-energy future. The company that owned it was valued at more than $1.5 billion, and its shares publicly traded. The factory, the first of several planned in the state, was intended to employ hundreds of workers and create new demand for the state’s timber industry.
But on a hot, sunny afternoon in October, the factory is a dead zone. Long weeds have sprouted up around an empty parking lot. No workers are operating any machinery. The plant hasn’t produced any biocrude in close to two years. Days before, a big chunk of the facility was sold for $2.1 million. Another piece was unloaded weeks before that for $1.6 million. The plant, a former paper mill, had cost more than $215 million to buy and convert to green energy production.
The factory was run by a company called KiOR, which was once a symbol of the promise of the next generation of biofuels and the role that Silicon Valley and government could play in incubating clean-energy technology. Its soaring ambitions—and hype—largely stemmed from the imprimatur of Silicon Valley venture capitalist Vinod Khosla and the millions he had invested in it. All told, the company spent more than $600 million. In its brief time in operation, it generated $2.3 million in revenue; when it filed for bankruptcy it listed assets of $58.3 million.
At one time KiOR was an important company for Khosla, whom Fortune called “the most successful venture capitalist of all time” in 2000. He’s a billionaire who moves in rarefied circles; he hosted a dinner for President Barack Obama at his home a few years ago. In 2004 Khosla left the elite venture capital firm Kleiner Perkins Caufield & Byers. He launched Khosla Ventures partly because he wanted to make an outsize bet on clean technology. The next generation of biofuels, made from plants and biowaste (so-called cellulosic materials), which have lower carbon emissions than oil, were a particular passion. Khosla invested hundreds of millions of dollars in about a dozen biofuel and biochemical companies.
His ambitions were audacious. Khosla declared “a war on oil.” As he wrote in 2006, “I believe we can replace most of our gasoline needs in 25 years with biomass.” He dismissed incumbent energy companies in a 2007 interview as not investing heavily in biofuels because they weren’t “used to innovation and the rate of innovation we are likely to see in this business.”
KiOR was a crown jewel in Khosla’s biofuel portfolio. Khosla Ventures held 75% of its voting shares at one point and wagered nearly $160 million, much of it Khosla’s own money. He attracted a constellation of names. Former Secretary of State Condoleezza Rice joined KiOR’s board. Later, tech magnate Bill Gates, who has invested in Khosla’s funds and shares his interest in energy tech, committed millions. Former U.K. Prime Minister Tony Blair joined Khosla Ventures as a senior adviser in 2010, partly to counsel clean-tech startups.
Khosla’s ambitions were audacious. “We can replace most of our gasoline needs in 25 years with biomass,” he asserted.
Yet only 2½ years after a gala groundbreaking, KiOR’s Mississippi facility, riddled with problems, stopped producing biofuels. Eleven months after that, in late 2014, the company filed for bankruptcy.
Unlike most failed startups, KiOR hasn’t just shut its doors and disappeared into oblivion. Today recriminations, investigations, and litigation continue to surround it. The Securities and Exchange Commission has been examining whether the company made false statements, including on a critical point: the yield of its biofuel (the amount that can be made per ton of wood chips). Two KiOR executives and Khosla himself are also facing a class action suit alleging that company executives misled investors about production volumes and yield.
The state of Mississippi is also suing Khosla and key KiOR executives on similar grounds, claiming they hoodwinked the state to obtain a $75 million loan. The case provides a striking image. One of the poorest states in the country, which received only $6 million of its money back before KiOR went bankrupt, is seeking to collect from a Silicon Valley billionaire. In the suit, Mississippi Attorney General Jim Hood described KiOR as “one of the largest frauds ever perpetrated on the State of Mississippi.” The state development agency called for KiOR to be liquidated and accused the company of manipulating the bankruptcy to give Khosla a better deal on the assets. KiOR countered by accusing the state of a “scorched-earth litigation strategy” that scared away potential buyers of the Columbus plant.
This past summer a judge approved KiOR’s Chapter 11 plan: The company sold itself to an affiliate of Khosla’s, trading $15 million in debt for equity, and receiving $29 million in exit financing. The plan enables KiOR, which had about 70 employees in its Texas headquarters as of the summer, to conduct research for at least a year and figure out how to make its biofuels at a larger scale. KiOR’s supporters say the core tech is still valuable and could someday be ready to commercialize.
Today many of the other biofuel startups that Khosla Ventures backed over the years have either shut down, been sold for small sums, diversified, or migrated to making bio-based materials and specialty chemicals. Khosla has argued that the stakes for the planet are so high when it comes to energy that the potential benefits easily justify the costs of flops (which, as with all venture investments, are numerous). In one blog post in 2014, he wrote that “to get to the energy-independent future we need, we must continue to try and sometimes fail, but the consequence for not trying is guaranteed failure. We will keep accepting intelligent and selective failure.”
Khosla and KiOR declined multiple requests for interviews for this article, but court filings for both deny any fraud or deception. In the past, Khosla has acknowledged that at least half his green-energy businesses would fail.
How KiOR crashed so disastrously is a complex saga. The company struggled—and ultimately failed—to deliver the yields it needed. It faced constant technical problems at its factory. It was buffeted by outside forces, including a worsening environment for green-tech financing and a drastic fall in oil prices. All of that was enough to kill off plenty of biofuel startups.
But KiOR’s woes were compounded by hubris and overweening ambition. It has become perhaps Silicon Valley’s favorite cliché to rhapsodize about the virtues of “failing fast,” and KiOR certainly accomplished that. But what is the practical lesson of that failure—that inventing new biofuels is even harder than it looks? Or that fast-moving venture capital investors are ill-suited to tackle such a technically demanding, time-consuming endeavor?
The Magnolia State
Mississippi Attorney General Hood is determined to hold Khosla and KiOR’s executives accountable. In an interview in his office on the 12th floor of the Walter Sillers State Office Building in Jackson, he says, “You can’t come into a poor state like this and gamble on $75 million of our money, and expect to go into bankruptcy court and walk away.”
A native of Chickasaw County in northeastern Mississippi, Hood has been a prosecutor for 25 years. He initiated the KiOR suit after lawyers obtained close to 400,000 pages of documents through the company’s bankruptcy case. Those documents, he says, show that the individuals involved “knew when they brought this proposal to us that their technology was incapable of producing the product that they claimed it would.”
Hood isn’t the first Mississippi attorney general to wage a populist campaign against an out-of-state company. One predecessor was instrumental in extracting a $246 billion settlement from tobacco companies in the 1990s. During his tenure, Hood has taken on corporations ranging from MCI WorldCom to Google GOOG and Microsoft MSFT . He sometimes retains contingency-fee law firms to sue companies, as he did in the KiOR suit, which saves the state money. Hood has received hundreds of thousands in campaign contributions from plaintiffs lawyers who handle cases for the state.
Seven floors above Hood’s office is the governor’s office. It was there that the state’s deal with KiOR was struck. On July 1, 2010, a group of 10 representing KiOR, the state, and Khosla Ventures met to discuss a loan. Hood says he happened to bump into the group on their way in.
The two most powerful men in the room were Khosla and Haley Barbour, then governor, who combined Washington savvy (he co-founded a major lobbying firm), a national presence (he is the former head of the Republican National Committee), and the molasses drawl of his hometown, Yazoo City, Miss. The two were introduced in 2008 at Teddy Forstmann’s annual Aspen conference and had discussed bringing a high-tech “smart” window maker called View (then known as Soladigm) to the state.
Barbour was interested in attracting new jobs to the state, which routinely ranks at or near the bottom in economic development metrics. Clean energy was particularly appealing because it often brought support from the federal government. Barbour was impressed by the pitch made by Khosla and Fred Cannon, KiOR’s CEO. A few months later he led the way, and Mississippi’s legislature passed a law that cleared the path for a $75 million loan. In return, KiOR promised to spend $500 million and build three biofuel factories in the state, use Mississippi-grown timber, and create more than a thousand direct and indirect jobs by the end of 2015.
“A Very Visionary Man”
The biofuel that would face its crucible in Mississippi actually began its journey in Holland. In 2005, Dutch chemical engineer Paul O’Connor left his job at chemical giant Albemarle to launch a startup called BIOeCON. His plan was to apply a process used in the oil industry to make biofuels.
The technology, called “catalytic cracking,” uses a catalyst to break down biomass, such as wood chips. Many companies and researchers at the time were—and still are—searching for a way to produce fuels using plant waste rather than inefficient corn or other sources that can put stress on food supplies.
Around that time the venture capitalists of Silicon Valley, who had made billions of dollars funding computing and Internet startups, were looking for the next big thing. In 2006, when O’Connor began seeking financing, Silicon Valley’s wallet was wide open. Khosla Ventures was one of the most aggressive in its focus on clean tech.
O’Connor managed to secure a meeting with another Khosla Ventures partner and flew to California. During the meeting, he asked if he could get the ear of the name partner himself and was offered five minutes with Khosla the next morning.
In his quick pitch that morning O’Connor described the catalytic cracking process. Khosla happened to be looking for something like that to round out his firm’s biofuels portfolio. He was interested.
After months of due diligence, the two sides agreed on a deal, and in 2007 they launched KiOR. (The name doesn’t mean anything; it was just one of many four-letter names Khosla Ventures had trademarked.) Soon after, they recruited O’Connor’s former boss at Albemarle, Cannon, to lead KiOR. Samir Kaul, a Khosla Ventures colleague, joined KiOR’s board and became Khosla’s partner on the project.
One of the most fateful decisions occurred even before the company was founded. O’Connor was considering licensing his technology to a big oil company. But Khosla—who can be almost as brusque and certain in his conclusions as he is intelligent—disagreed, according to O’Connor. He argued that there was no reason to solicit VC funding if O’Connor planned to sell the technology.
Khosla’s ambition was much bigger. He wanted to make KiOR a producer—a biofuel version of Exxon. That would require massive capital expenditures and huge teams with extensive technical know-how. O’Connor agreed, and says he relinquished a research and development agreement he had struck with Petrobras and stopped pursuing technical discussions with Chevron.
The stakes had been raised before the game even began. “Vinod is a very visionary man,” says O’Connor. “Maybe too visionary.”
Khosla has long played very active roles in companies and relentlessly challenged them. He explained his strategy in a 2011 Harvard Business School case study that examined his biofuel investments. Khosla said he considers what he does not so much venture capital as “venture assistance.”
Khosla and Kaul also helped the company bring in managers in a process that Khosla calls “gene-pool engineering,” and still uses today. The idea is that the initial team should know the industry well enough to be able to operate within it but also be outside of the industry enough not to be constrained by its traditional thinking. The group should be able to thrive in “guided chaos.”
Different parties disagree about which side was responsible—Khosla Ventures or O’Connor and the CEO—but most agree that KiOR made poor hiring decisions as it staffed up. The result was a relative preponderance of lab researchers with Ph.D.s and a dearth of people with technical, operational experience running energy facilities. The lack of people with real operational experience “hurt KiOR a lot,” says O’Connor.
The venture capitalists and the executives took another step that would put pressure on the company: selling KiOR stock to the public. That would subject the company to the scrutiny and burdens of the markets and outside shareholders—before it had ever sold a single drop of fuel.
Signs of tension were already emerging inside the company. One KiOR executive allegedly told Khosla in 2010 that the company wasn’t ready for a public offering because its technology was not yet fully developed, according to filings in the shareholder litigation.
Many strategies the venture capitalists employed at KiOR are commonplace in Silicon Valley. They tend to work in the computing and Internet realms, where top engineers are easy to find and startups often grow and launch products incredibly quickly. These approaches can allow tiny software startups, such as Uber and Skype, to upend massive industries. But inventing and producing a new fuel is more complex by orders of magnitude, the time frame is dramatically longer, and the budgets are astronomically larger.
In late 2010 and early 2011, KiOR’s future seemed to glow. On May 12, 2011, the company held a groundbreaking ceremony at the Columbus facility. Barbour and local politicians donned white construction helmets and posed for photos with shovels. A local elementary school choir performed. The governor called the factory a “game changer for our country.” He described KiOR’s process as “almost like making gold out of straw,” according to the local newspaper.
Three days later came a less public event, but one that would cost a potentially huge support: KiOR suspended an application for $1 billion in federal loan guarantees. The reasons for the company’s decision are murky, but it meant KiOR needed to ramp up its fundraising.
In light of that, its ability to raise $150 million in an IPO, which happened the next month, offered some relief. But there were at least a few hints that the market was skeptical of Khosla’s baby: KiOR had hoped to raise more and had to trim back when investors showed limited enthusiasm. Still, the company was on its way. Now all KiOR had to do was develop and sell commercially viable biofuel.
The Rise and Fall of a Green Factory
KiOR’s IPO documents stated that the company had crossed a crucial threshold: KiOR said it was able to turn a ton of biomass (technically a “bone-dry ton,” or BDT) into 67 gallons of fuel. At that yield, the company would eventually be able to produce its fuel for $1.80 a gallon and compete with traditional gasoline. During investor calls in 2012, CEO Cannon said KiOR expected to achieve 72 gallons at a larger commercial production facility in the future and would aim to reach 92.
The problem was that the company wasn’t regularly producing anything close to 67, much less 92. In the summer of 2011, after the company’s IPO, its new chief operating officer, Bill Coates, told management that it appeared that the company had grossly overstated its yields, according to a KiOR consultant who spoke with him and was quoted in the state’s court papers.
Coates’s worries fell on deaf ears, according to the consultant. The conflict escalated, and the COO accused his fellow executives of malfeasance, claiming they had “cooked the books.” Coates told the executives he was not “going to be part of this scam.” A short time later he resigned, signed a nondisclosure agreement, and left with a year’s salary and other benefits.
The fight wasn’t visible to the outside world. Indeed, at that point in 2011, KiOR’s stock was trading high, and two other Khosla Ventures–backed biofuel companies also went public. Another two filed for IPOs before the end of the year (but later pulled their offerings). Khosla proclaimed in a speech at the time that his fund’s biofuel portfolio contained about $1 billion in “liquid profits.”
November 2012 brought more good news when KiOR announced it had made its first biofuel at the Columbus factory. That caused Cannon—the soft-spoken, deferential CEO—to declare it “the world’s first cellulosic gasoline and diesel fuel products.” In an earnings call that day, CFO John Karnes said KiOR would be able to meet its target of selling between 500,000 and 1 million gallons of fuel before the end of the fiscal year.
That’s not at all how it played out. The facility was bedeviled by production problems. The conveyor mechanism that delivered wood chips was often on the fritz. Cleaning systems routinely jammed with a tarlike substance. The company spent tens of millions of dollars more than it had expected, and its researcher-heavy staff couldn’t untangle the problems.
Five months later KiOR revealed that it hadn’t shipped any biofuel in 2012. The facility hadn’t yet reached a point of steady production.
An odd pattern soon emerged: gloomy news paired with sunny projections. In KiOR’s earnings call in March 2013, Cannon said the company had only just shipped its first commercial fuel the day before. He was “disappointed” that the factory had missed its 2012 target, but KiOR had encountered “unexpected startup issues unrelated to our technology,” which they had since “overcome.”
The company continued issuing bullish estimates. On May 9, 2013, Cannon said the factory was running as expected and cited a production rate that put it on track to hit its 2013 target of making between 3 million and 5 million gallons.
But KiOR was still not shipping fuel regularly. And even when it managed to do so, KiOR’s yields were only a fraction of 67 BDT. Its two plants were routinely producing between 20 and 40 gallons per ton.
Concerns were rising internally. An experienced chemical engineer named Mark Ross joined KiOR in Texas in the summer of 2013 to help fix its endless technical snafus. He visited the Columbus factory multiple times, and when he found that the yields were closer to 22 gallons per BDT, he quickly grew worried, according to a sworn statement he later filed as part of a whistleblower complaint he submitted to the Department of Labor.
Ross asked questions about the yield discrepancy so frequently that he earned the nickname “Dr. Doom.” At one point, according to his sworn declaration, he told a superior that the public yield data was so different from the actual data that he believed the company was “lying and cheating the investors.” Ross was fired in January 2014 because of what KiOR described as a safety violation. (Ross entered an area of the factory without a permit.) His whistleblower complaint was later dismissed.
KiOR’s yields were so much lower than what execs said that one employee accused them of “lying and cheating the investors.”
KiOR’s co-founder, O’Connor, and a senior scientist, Dennis Stamires, also began to worry about the yield discrepancies, according to Mississippi’s pleadings. O’Connor departed as a full-time employee early on but says he sent a series of emails to Khosla expressing his fears. In early 2014, Khosla brought O’Connor back as a consultant to review the technology and production.
Even as KiOR struggled to achieve effective production at its Columbus factory, it was also planning a second, larger factory in Natchez, Miss. That factory would also cost hundreds of millions of dollars to build, and the company could get financing for it only if it showed that the technology was working well at Columbus.
It wasn’t. KiOR had once again tripped up, as it revealed in an earnings call in August 2013. Its output was 75% below its minimum target.
Finally, near the end of 2013, the company began unraveling. Rice quietly resigned from the board, issuing a statement that she had too many commitments. KiOR disclosed that it had been hit with a securities class-action lawsuit. The company’s CFO resigned a few months later. Its stock cratered by 90% from its IPO price. In March 2014, KiOR revealed it had been subpoenaed by the SEC and had to disclose that it didn’t know if it could continue as a “going concern.”
Khosla tried to prop up the company, announcing $100 million worth of equity commitments (including a reassuring contribution from Bill Gates).
But it wasn’t enough to fix the problems at Columbus, and at the beginning of 2014, KiOR closed the plant. Cannon shared the bad news with investors: “We have learned from a year of operating the Columbus facility that additional work is required to bring Columbus from its current performance … with a yield in the low 30s in terms of gallons per ton.”
With a shuttered plant, no revenue from fuel sales coming in, and loan payments looming, the company ran out of money and time. It tried and failed to find a buyer. Then, in November 2014, KiOR filed for bankruptcy.
No biofuel startups have managed to produce a next-generation biofuel at commercial scale in the U.S. The capital costs have simply been too high, and the time lines too long. The collapse in oil prices has only made the challenge more daunting.
The list of biofuel startups backed by Khosla Ventures now reads like a catalogue of disappointments, vaporizing hundreds of millions of the firm’s money, along with millions in government funds. Range Fuels built a cellulosic ethanol plant using Khosla funding and millions from the DOE and the USDA. The company shut its factory in 2011. Another half-dozen biofuel startups backed by Khosla Ventures have failed to deliver significant results.
Two of the ones that looked successful early on, Gevo GEVO and Amyris AMRS , both went public. But today their shares are trading at a fraction of their initial value. In 2012, Amyris decided to move away from biofuels and focus on higher-margin specialty products. Other biofuel companies have done the same. One Khosla-backed biofuel company that has shown promise is LanzaTech, which turns industrial waste gases into biofuels and biochemicals. It has a demonstration-scale plant in China and plans for a commercial facility there.
Fewer than 2 million gallons of cellulosic biofuel were produced in the U.S. in 2015, according to the EPA. That’s a small fraction of the 3 billion gallons the agency had predicted in 2007 would be produced this year. Only four cellulosic fuel plants have been built and are in the commissioning phase in the U.S., according to Bloomberg New Energy Finance. All were built by large companies, including DuPont DD and INEOS. Looking at those numbers, KiOR accomplished something remarkable: It made close to a million gallons of cellulosic fuel in a year, something almost no other company has done in the U.S.
The drama will continue to play out. Mississippi’s litigation against Khosla and the KiOR executives is now waiting for a judge to decide whether it will be fought in the federal bankruptcy court in Delaware (where the defendants would like it to occur) or in a district Mississippi court, where Hood wants it. “The deal was done here. Everything was done here,” the attorney general says.
Once the litigation clears, O’Connor and others dream that KiOR’s technology can eventually succeed. Indeed, Khosla’s willingness to purchase the remnants of the company suggests he still believes in the technology. Can he still engineer a bio-Exxon in the future? It’s a long shot. But if the size of the failure equals the magnitude of the educational payoff, then KiOR will have been a very valuable lesson indeed.
A version of this article appears in the December 15, 2015 issue of Fortune.