Nevertheless, CCTER within the DOE family could be to climate change mitigation what DARPA is to the military.

A second option might be a special Federally Funded R&D Center (FFRDC) such as the Air Force's Aerospace Corporation. Such a corporation could be created to provide more flexibility and more insulation from the requirements imposed than if CCTER were organized in DOE. This option should be carefully considered. One possible variation on this theme is the NASA Institute for Advanced Concepts (NIAC). It was set up administratively outside of NASA for the purpose of functioning as an independent source of revolutionary aeronautical and space concepts that could dramatically influence how NASA develops and conducts its missions.

The third option is a private not-for-profit corporation. An example is RAND Corporation set up originally after World War II as a think tank for the DOD, but now does work for many agencies. The difference is that CCTER would be a corporation that funds R&D using both private and public sector funds. Several NSF centers operate this way, for example, the Aspen Center for Physics is a not for profit corporation funded by NSF and others. Under this third option, CCTER would have a board of directors with representatives from both DOE and the private sector sponsors. It could have considerable insulation from DOE politics and bureaucracy as well as from private sector pressures. It could be very flexible, and it should be able to attract top talent. For these reasons and because of the need to manage private and public sector resources productively, we conclude this is our preferred option. Taking maximum advantage of private sector intellectual contributions is a very important in-kind asset that a private not-for-profit corporation can generate more readily than other organizational options.

4. How much government money is required?

The answer to this question is a judgment call. We believe that CCTER should operate in the following manner. The first year it should solicit proposals from which the most promising would be selected for support. Obviously, some exploratory research may require more money for proof of concept than other ideas. By their nature, some may require several millions of dollars a year to test while others may require only a few hundred thousand. This is clear from an examination of the examples in Box 1.1. It may be useful to divide the funding so that some expensive projects can be examined each year. Of course, it is probable that most ideas that show promise after CCTER seed money funding will require more resources to fully demonstrate and initiate deployment. This maturation investment could come from either DOE or other CCTP agencies or the private sector, and one vital CCTER function would be to fully encourage needed follow on support.

We suggest, therefore, two categories of proposed research. Category 1 projects would include paper studies or small laboratory scale proof-of-concept experiments with annual costs typically in the range of $100,000 to $500,000 per project. Category 2 projects would test the engineering and cost potential for ideas that have already been vetted at the paper study or bench-top scale. Annual funding levels for these contracts might average

in the range of $500,000 to $1,000,000. In general, the Exploratory Research contracts would be for two or three years with extension possible but not common, although successful Category 1 projects could submit Category 2 proposals.

Assuming funding for 20 to 30 ideas per year with equal number of each category and 3 year funding, steady state expenditures for CCTER could be in the range of $35 to $50 million/y. To this must be added the costs of operation including organizing the peer review and evaluation process, and the cost of maintaining contacts with top talent and institutions around the world that may provide introductions to people with revolutionary new ideas and insights. These extra costs may be in the range of 10 to 20% of the contract awards. At steady state, the cost would be shared between the Federal government and private sector contributors. If it were on a 50/50 basis, the Federal cost would be in the range of $19 to $30 million per year. Conservatively we believe the order of $25 to 45 million/y of Federal money is needed at steady state because it is likely that private sector support will be less than 50/50, at least initially.

Of course, the CCTER should start at a much lower level until the concept and procedures are fully worked out and tested. No doubt, there will be some growing pains.

We suggest starting at $5 million per year for the first year, funding primarily Category 1 proposals, and ramping up from there to the steady state level in 5 years.

This Federal funding for CCTER is very small compared to the magnitude of the overall CCTP portfolio that is in the $3 billion per year range, but we believe this small flexible seed money type of investment will have payback far in excess of the investment.

5. What process should be used to select projects for funding and how should CCTER be evaluated?

Proposals would be solicited very broadly including from universities, commercial organizations, national laboratories, and even foreign organizations. Panels would be set up to evaluate the proposals, and these would include people from DOE and other agencies and from private sector donors as well as from the technical community at large.

The membership of the panels would be changed periodically.

Criteria for judging each proposal should include: 1) the potential impact of the proposed idea on climate change mitigation assuming realistic optimism for all relevant factors including cost, 2) the probability of success, 3) technical and scientific merit and risk, 4) the fully loaded project cost, and 5) potential confounding issues such as environmental impact, safety, infrastructure, and geography. The division of 1)*2) by 4) might give a crude estimate of return on investment. The portfolio of investments could also be balanced in terms of probability of success to provide some long shots and some medium-shots. Votes on these criteria could be measured on a median-basis so a few naysayers or zealots on the panels will not skew the results too badly.

Progress by funded projects should be evaluated annually. We suggest Category 1 projects be evaluated by CCTER management. Category 2 projects should be evaluated by peer review. This way mid-course corrections or even cancellation can be invoked to avoid waste.

CCTER itself should be evaluated periodically to assure the mission is being pursued effectively, and to evaluate whether the investment is yielding adequate return. We suggest that this evaluation be done by the National Research Council (NRC) with a committee composed of people with different backgrounds with no direct conflicts of interest. The measure of success is the number of unique ideas that are judged to have potential for making a big difference if the cost is right. This NRC report would go to DOE, associated sister agencies, other sponsors, Congress, and the public.

6. How could CCTER be initiated?

The first step is to generate enthusiasm for the idea of CCTER. It should be done within DOE, in the Congress and among the general public. The idea should be thoroughly vetted including in the private sector and academia. Assuming the vetting results are generally positive, a decision should be made between the three options of Section 3.

Assuming option 3 is chosen (or even option 2) a not-for-profit corporation should be set up. Money for this activity might be found from one or more foundations. We note that the formation of RAND was funded by a grant from the Ford Foundation. The corporation could then choose a CEO, appoint a board of directors and organize the solicitation for proposals. Simultaneously, work would go on with DOE CCTP, other agencies, OMB and Congress to propose, authorize and appropriate the first year of funding. With the arrival of funding, CCTER is operational.

ful.

Chairman TOM DAVIS. Thank you very much. That is very help

Dr. Van Atta, welcome and thank you.

STATEMENT OF RICHARD VAN ATTA

Mr. VAN ATTA. I am not an energy specialist. My background is Defense and Defense research. I spent a fair amount of my career looking at emerging technologies and how they are made to emerge, and I teach a class at Georgetown on emerging technologies and security, and I emphasize the fact that emerging technologies are made to emerge. The question is the processes and the means by which you do that.

DARPA is a unique example of an entity that was created with that purpose in mind, and I think it is important to look at it in terms of why it succeeded and what made it succeed. In my testimony, which I will read portions of here, I emphasize that the research that DARPA does is unique and different, and is purposefully so. The organization, itself, is designed explicitly to allow it to do this unique and different type of research, and it has cultural features within its organization and management style that allow it to do that.

In the testimony I talk about the DARPA model and I also ask the question of which DARPA model, because DARPA has done many things in many different ways. It has been adaptive. It is very malleable. One cannot just say there is a DARPA and that we are going to take that and implant it some place else. You have to understand what it took to make it do what it could do and why it was able to change in those very effective ways. So it evolved over time and it has many successes, and those successes, in fact, were different because they were dealing with different problems. We have to understand the way in which those successes were made and what it took to make those successful, and I will talk about a couple of examples of that.

DARPA's program managers are the core. They are, in fact, almost individual entrepreneurs. They are encouraged to challenge existing approaches. In the case of Defense, for war fighting and to seek results rather than just explore ideas. In addition to supporting technology and the components of the technology development, DARPA has also funded integration of large-scale systems demonstrations to look into what we would call disruptive capabili

ties.

There is a high-risk, high-payoff motif for DARPA that is a set of organizational and operational characteristics that include its relatively small size, its lean, non-bureaucratic structure, its focus on potentially change-state technologies, its highly flexible and adaptive research programs, but what is most important at the outset is that, in contrast to the existing Defense research environment, ARPA was manifestly different. It did not have labs. It does not focus on existing requirements. It is separate from any operational organization elements. What is explicit is that its charter is to be different so it could do fundamentally different things that

So when one looks at an energy ARPA or climate change ARPA, the question is what are the things that it is trying to do that are different and how do you set up an organization to do that.

DARPA was established as a research and development organization to assure the United States maintained the lead in the stateof-the-art technology for military requirements and prevent technology surprise. As one then looks at the characteristics of how it did that, first of all it was independent of other organizations.

Second of all, it is lean and agile. It was risk-taking and tolerant of failure, open to learning. You have to have a specific kind of research environment and organizational structure and a way in which your link to the rest of the organization will allow you to do that.

The program managers are, in fact, the technical champions who conceive their own programs and have to then sell those programs within the DARPA environment. The coin of the realm in DARPA is promising ideas. Gaining notion is not that the idea is well proven, but that it has high prospects for making a difference on the problem they are trying to solve. So you have to have an organization and culture that focuses on those kinds of innovations and those kind of directions.

In my testimony I talk a lot about DARPA's successes, and I don't have time to go into those here, but I will give you some key what I consider to be elements of that success.

First of all, focus on creating surprise, creating difference, not avoiding them.

Second, build what I call communities of change state advocates. One of the key things that is unique to DARPA is it doesn't create and do its own research, it incentivizes and creates a community of people to do that. If one talks about the current structure of DOE in the national labs, they do their own work with their own capabilities within their own operations. What DARPA did is it found the people who could do that. It developed the community. It found the new ideas out there and brought them together in a coherent manner.

The third element is to find challenges, develop solution concepts, and then demonstrate them. We can show examples of that in my testimony.

Finally, I would say if one were to ask the question what were the key things about climate change that relate to DARPA and the DARPA model, the first thing I would say is you have to understand the imperative that drove the creation of DARPA in terms of national security, the Sputnik issue, and ask the question: do we have the same imperative and understanding of imperative to make an ARPA-like organization work elsewhere?

You also have to have the understanding that it will work because of the protection, oversight, and interest of the Secretary of Defense and even the President to make it happen. Without that, just naming something ARPA will not solve your problem for you. Finally, I would say you need to deal with not only leadership support but the issue of congressional oversight. ARPA has benefited from the fact it has a simple oversight structure, it is not being managed by multiple congressional committees simultaneously, and with that kind of multiple meddling you are not going