Clean-Energy Push Takes Off

“Nimble, lean, flat”—that is Arun Majumdar’s vision for the Department of Energy’s new Advanced Research Projects Agency-Energy (ARPA-E) that he has led for barely two months.

In a sprintlike 20-minute interview with C&EN, Majumdar bounds from subject to subject, outlining the future for his tiny agency with the huge charge of finding, funding, and guiding development of high-risk transformational technologies that can drive a clean-energy revolution.

Energy Secretary Steven Chu is watching Majumdar and the agency closely. From his first days in office, Chu has stressed his desire to light a fire under DOE-funded research, bypass the agency’s stovepipes of program-based research, and bring raw, out-of-the-box energy ideas to the doorstep of commercialization. For a model of transformative research, Chu points to his industry-funded research at Bell Laboratories and elsewhere that led to his Nobel Prize. He envisions ARPA-E as an incubator for the energy-based Industrial Revolution he seeks (C&EN Online, March 2, 2009).

Such transformations carry high risk. And creating a fluid environment to encourage freedom to explore ideas while still following government grant policies is rife with contradictions. However, the potential pitfalls have not dampened the excitement for and interest in the new agency by energy-related scientists at DOE and elsewhere: In the first round of ARPA-E grants held last year, some 3,700 applicants chased $151 million in funds, but only 1% were successful (C&EN, Nov. 2, 2009, page 9).

The concept of ARPA-E grew from recommendations in a 2006 National Academies report in part authored by Chu, who at that time headed Lawrence Berkeley National Laboratory (LBNL). Congress made the concept into law in 2007, modeling ARPA-E on the $3.1-billion-per-year Defense Department’s Defense Advanced Research Projects Agency (DARPA), a solution-based research management organization that develops raw technologies for military applications. However, Congress didn’t fund the agency; instead President Barack Obama did so through an infusion of stimulus funds from the American Recovery & Reinvestment Act of 2009.

Ironically, it was left to Chu to put flesh on ARPA-E’s bones. On Sept. 18, 2009, Obama announced the selection of Majumdar—who, like Chu, is another LBNL scientist—to run ARPA-E. Majumdar reports directly to Chu.

The close relationship between Chu, Majumdar, and LBNL is one of many concerns voiced by scientists who see all sorts of possible bias and favoritism in how research topic areas are chosen and how projects are assessed and selected under the new agency. Some also worry that the requirement that grant winners match 20–50% of ARPA-E funds puts companies in the driver’s seat because universities and government labs cannot scrape up the money.

Such fears should be expected, these critics note, because of the total power the small agency has over fund allocation. But the critics also note the great potential of the agency to do good and to change the energy world, as well as to reward some garage tinkerer or themselves with a multi-million-dollar grant.

ARPA-E’s forerunner, DARPA, grew out of a DOD research program founded shortly after and in response to the Soviet Union’s launch of Sputnik in 1957. It now has some 240 staff; about half are program managers and other scientists who work for three to five years at DARPA and then return to academia, business, or from wherever they came, DARPA spokeswoman Johanna Jones says. These program officers are key to keeping the program successful.

DARPA “is a very fluid, very flexible organization,” Jones says. “Our last director was here for eight years, and that was a record.” She adds that “our primary goal is to save the lives of men and women in uniform. We meet with the guys in the Pentagon. They tell us what keeps them awake at night, and then we ask how we can solve that problem, and our people run and do it. We only have a set amount of time.”

DARPA works on “hundreds of projects” at once, Jones says, and the project managers closely monitor them every year or 18 months and decide whether funding should be extended or ended, she says.

Mirroring DARPA, Majumdar explains, “ARPA-E’s program managers rotate and serve a maximum of four years. That way, we get the best and the brightest, the all-stars,” he says. “That is true for me, too. I will be here for a few years, and then I go. It is a constant shuffle. We are always looking for new ideas, fresh approaches. We have to be nimble, agile, and adaptive to what the conditions are. We are not going to have someone here for 20 years. We need new people: Youth is our future.”

The agency now has but three program managers: David Danielson, who holds a Massachusetts Institute of Technology materials science doctorate and was, until joining ARPA-E this summer, employed by a Cambridge, Mass., venture capital firm; Mark Hartney, a chemical engineer with degrees from the University of California, Berkeley, and MIT and a 25-year veteran in technology development, including a stint at DARPA; and Eric J. Toone, a professor of chemistry and biochemistry at Duke University who specializes in physical organic and bioorganic chemistry and studied at the University of Toronto and Harvard University.

“Right now, we have a small staff, seven or eight, but we are interviewing a lot of people on the program side. We want growth, but I want to take our time getting there and get the best people,” Majumdar tells C&EN.

According to the law that created ARPA-E, the agency is to have 70–120 program managers, but Majumdar says those numbers will come with a much larger budget. Over the next six months, he hopes to recruit around 12–15 program managers.

“These are deeply technical people who are going to be engaged with each project we fund,” Majumdar says of the program officers. Their goal, he notes, “is not to micromanage” projects, but to help investigators overcome technical obstacles by, for example, recommending researchers who may be able to assist on a sticking point or to point out a relevant research paper.

Program officers may be critical in keeping projects moving forward. “All these projects have milestones and they have to be met in a certain time. If you don’t meet certain goals, your money is going to be pulled,” Majumdar says.

ARPA-E is now buzzing with activity, in stark contrast to its early days. ARPA-E was but a concept and had not received funding until April 2009, when Obama flooded it with $400 million in stimulus money.

With personnel borrowed from other DOE programs, ARPA-E opened for business, and on April 27, 2009, it issued its first call for concept papers—with a short June 2 deadline. The little agency’s popularity with the scientific community was overwhelming. It received some 3,700 concept papers, offered by scientists from universities, companies, government labs, and other institutions.

Quickly, the agency began hiring its first program managers and searching for reviewers for the papers. Operations stalled, however, and complaints and rumors began surfacing in the scientific community over the status of projects and difficulties in finding knowledgeable reviewers who didn’t also want a grant or who wouldn’t profit intellectually or financially from having a chance to see a competitor’s ideas.

Eventually, some 500 reviewers were selected, and they and ARPA-E staff began culling the submissions. They examined 334 full proposals and selected 37 winners. Most of this happened without Majumdar at the helm; he wasn’t confirmed by the Senate until just a week before the winners were announced on Oct. 26, 2009.

The average grant from the $151 million allocation for this round was $4 million per project; grants ranged from $500,000 for a Lehigh University project to capture carbon dioxide through an electric field swing adsorption technology to $9.15 million for Foro Energy’s plan to use a thermal-mechanical technology to drill through basement rocks in a search for geothermal energy.

The agency is now negotiating final agreements and milestones for projects selected in this first round, which it expects to complete this month. But the race is on: Under the terms of the stimulus package, all $400 million is to be deployed by Oct. 1. Also, ARPA-E issued another call for proposals last month, this one for $100 million in three specific areas: biology-based and nonbiomass or petroleum, liquid transportation fuels; advanced carbon-capture technologies; and transportation storage batteries.

When asked what the new agency’s metrics of success are so far, Majumdar says with a laugh that processing 3,700 applications is clearly one.

“The first round was wide open. We wanted the best ideas no matter the topic,” Majumdar says. “It was good because you get to know what people are thinking, but you have huge logistical issues internally. This second round is much more specific and is more typical as we are getting into a much more steady-state operation.” But, he adds: “once in a while, though, we might do a more open request. You never know what is out there, and if you are too focused, you might miss what some kid in a garage may be thinking.”

However, some have expressed concerns about how ARPA-E works.

When the first list of 37 winners came out, several scientists told C&EN they were disappointed by the number of large companies—such as DuPont, Archer Daniels Midland, ConocoPhillips, Albemarle, and Babcock & Wilcox—that led or were part of successful consortia.

“I thought this was supposed to be for small start-ups,” says one scientist who doesn’t want to be named because he may apply for an ARPA-E grant in the future. “These companies don’t need the money.”

More than 60% of the projects are led by companies, Majumdar acknowledges. But he counters that 43% of the grants went to small businesses, and only 19% went to large companies. “We are not going to limit ourselves,” he stresses. “If you have good ideas, we will fund you.”

Criticism also was directed at the selection process itself. Between October and December 2009, ARPA-E held five workshops to help guide project calls. Their size was limited to around 80 attendees, and researchers were turned away. Members of the press were not allowed to attend or observe the discussions. To some, the workshops—which helped set selection guidelines—lacked transparency.

“You obviously can’t have 300 people taking part in a conversation,” acknowledges one researcher and potential ARPA-E grantee, who asked not to be named. “But lots of people were turned away. Although we know ARPA-E is a work in progress, you get the impression it is led by the program directors, Chu, and Majumdar without a lot of other input.”

Secrecy raises other concerns. In its announcement of first-round selections, DOE provided little information for some of the projects. For Foro Energy, the company winning the largest award, it provided almost nothing—just a sketch of Foro’s research goals and its Littleton, Colo., location. The company has no phone listing at Littleton, and its website (foroenergy.com) says: “Foro Energy is a drilling technology company backed by North Bridge Venture Partners and CMEA Capital. Please direct inquiries to Joel Moxley.”

In an e-mail response, Moxley said: “I can confirm that we are proud to be a Colorado-based company and were honored to be selected for award negotiations for a DOE ARPA-E grant. Other than that, we are stealth/quiet mode and not talking about what we are doing.”

When quizzed, DOE would only say, “Regarding your request to be put in touch with Foro Energy: Our contact at Foro has asked us to politely decline any media requests at this time and has asked that we do not give out their contact information.”

Some scientists, concerned about bias and the LBNL connection, say they fear that ARPA-E will become an insiders’ club, biased toward technologies and scientists known by the ARPA-E staff and advisers. They point to a recent ARPA-E call for liquid transportation fuels projects, arguing that the call reflects a bias for projects with a synthetic biology approach, similar to work done at the California lab.

Criticisms aside, the first-round grant winners appear eager to start.

“We want to use ARPA-E to get to the prototype stage and then to manufacturing,” says grantee Tom Kennedy, chief innovation officer of FloDesign Wind Turbine, an offshoot of FloDesign, a 20-person Wilbraham, Mass., aerospace technology engineering firm. The firm won a grant to pursue new wind-energy design.

Through its aerospace experience and venture capital support, the company developed the concept of a small, highly efficient wind turbine, Kennedy explains. Their wind turbine would be easier to move and install and much quieter and more adaptable than current huge wind turbines with 100-meter blade configurations, Kennedy says.

The company, he says, will use its $8.3 million, 30-month grant to overcome two “valleys of death” for funding: the first comes when a project transitions from basic research to prototype construction and testing, and the second is when it transitions from prototype to a manufacturable product.

In another typical ARPA-E project, Nalco, a water treatment manufacturer and process engineering firm, and Argonne National Laboratory received a $2.3 million grant for a 24-month project to capture CO₂ emissions from coal-fired power plants. The project uses carbonic anhydrase to capture CO₂ from a flue gas stream at almost 100% purity, the company says. If successful, the electrochemical process could significantly reduce the 30%-plus energy loss associated with existing capture technologies.

The project is based on a relationship forged over the years between Nalco and Argonne, says Seth W. Snyder, a section leader in Argonne’s Energy Systems Division.

“It is really a 50-50 effort,” Snyder continues. “We have the research expertise, and the company knows equipment and scale-up. We cannot see a way for us to do it without them, and they can’t do it without us.”

Nalco is putting up a 20% match to the DOE funds by buying equipment and hiring more personnel for itself and Argonne, says Cathy C. Doucette, Nalco’s global technology leader. Their joint goal, she says, is to move the technology from small bench-scale to a “robust bench level” sufficient to demonstrate the economics and ability to build a prototype.

With 11,000 employees worldwide, Nalco is based in Illinois, just 10 miles from Argonne. Their proximity and familiarity with one another played a big role in laying the project’s groundwork, Snyder and Doucette note.

Scientists underscore the importance of partners’ history and location. Without such a relationship, they note, matching funding can be hard for university and government labs to obtain, even if they are developing breakthrough technologies. As a result, many worry that ARPA-E projects may be dominated by industries, particularly large companies with deep pockets.

Looking to the future, Majumdar says he wants to see slow growth in funding as ARPA-E shifts from stimulus funds to congressional appropriations.

The verdict is still out on congressional support, but the scientists who spoke to C&EN—including those who voiced criticisms—want Majumdar and ARPA-E to succeed. Success may depend on what happens between now and Oct. 1, when the $400 million stimulus money runs out.