to be considered meaningful for the investment of funds, this is not the case if R&D is viewed as “buying information" for future potential options. The "benefit cost” method of evaluating options would exclude a variety of potentially very beneficial projects and technologies simply because we do not have enough information. When only sketchy information is available many projects would be excluded by a negative expected net present value, even though there exists a sizeable chance, say 30% or 40% of a positive net present value. This is illustrated in Fig. 1. Many unique R&D and technological opportunities may prove to be of tremendous economic value, yet have negative expected net present value. But they also promise a "chance” to be potential winners, though the uncertainty of today is illustrated by a wider spread of the likelihood of expected values. R&D, and information generated by R&D, gives us a chance to find out over five or ten years whether the proposed technology is economically viable or not. That is, R&D expenditures can identify whether distribution 1, 2, or 3 of Fig. 1 may hold by 1985. R&D narrows uncertainties, but “buy information” analysis is not a license to finance any long-term R&D project that comes up for consideration. Rather, its rigorous analysis should show convincingly that the expected reductions in uncertainty from the proposed research will indeed occur. To perform such evaluations rigorously is, however, quite difficult. 3. Unique R&D Programs: There are a few, major R&D technology initiatives which can and should be viewed differently. When a proposed technology promises a unique, otherwise unachievable, breakthrough conventional criteria of benefit cost analysis or even R&D to “buy information” criteria break down, or may not do justice to the proposed potential idea. This is the case when any “finite” set of existing capabilities cannot reproduce the promised capability. Examples that come easiest to mind are in the area of national security: nuclear explosive technology in World War II and thereafter; nuclear submarines; and global space observation systems of national and military resources. Such projects lack the “Archimedean property,” that a finite set of existing capabilities can provide the capabilities of the proposed new technology. SPS may have aspects of “non-Archimedean” properties; for no long-term, feasible, environmentally safe, economic energy options are known to me today that have been proven in principle to provide most if not all of mankind’s energy needs. Clearly, fossil fuels are finite and exhaustible. Nuclear fission power has serious environmental risks and problems remaining, and in a strict sense is not an inexhaustible energy source. Controlled nuclear fusion energy for generating electricity still is years away from showing the technical feasibility of producing net energy output. SPS technology provides net energy output and has been used extensively in space and on the ground for decades. The only questions with SPS are whether systems of this scale and size can be deployed in space safely and reliably, and at economically competitive costs. If SPS proves to be the only such long-term large- scale option with a promise of providing significant energy requirements for all of mankind, not for just a few nations, it may indeed promise benefits at economic costs that cannot be matched by any other reasonable combination of energy alternatives. Yet, I do not believe this to be a sound basis for SPS funding. SPS technology and funding has to be evaluated as buying information through research and development. This is illustrated in Fig. 2. Today, under a variety of specific technical and economic circumstances, SPS promises to be an economic system. Alternatively and with equal validity, if some critical economic, technical, and environmental issues cannot be satisfactorily resolved, SPS may prove to be uneconomic five or ten years
RkJQdWJsaXNoZXIy MTU5NjU0Mg==