obvious source of certain basic materials, because of the much lower energy requirements for transport off the lunar surface. However, against this advantage must be weighed the logistics costs of maintaining an operational base on or near the moon and bringing up from earth those elements (hydrogen, for example) which are required for operations and are not available on the moon. Manufacturing costs may be conveniently broken down into recurring and nonrecurring costs. Nonrecurring costs are the initial costs associated with, for example, setting up lunar and other space bases, acquiring the necessary equipment, and prototype development. Recurring costs are those associated with the actual manufacture of the sub elements and their assembly. These latter costs are subject to an exponential learning curve in which the cost of production of a single unit decreases as the number of units built increases. Unit costs may therefore be expressed in the form where Cn is the cost of the nth unit in a production run of N. For a learning curve corresponding to a 20% reduction in cost for every doubling of the production run (an 80% learning curve), P has the value P = .32. It is therefore clear from (1) that the cost of producing an SPS will depend critically on N, and it may furthermore be expected that the optimum manufacturing scenario will also depend on N, which represents the total thruput of material through the system. Another primary factor determining industrial costs is the productivity of labor in space. This factor is important not only because of the wage scale, even when allowing for the much higher anticipated scale in space, but also because of the costs of life support and the need for crew rotation which, in turn, are critically dependent on transportation costs. These costs would be much reduced if it were possible to postulate the availability of space habitats with closed ecological systems: indeed, the future development of space power systems may well be the catalyst leading to the realization of the concept of space colonization. Recent experiments in a simulated weightless environment (4) have provided some information on the potential productivity of man in space; however, this dominant cost parameter has been kept a variable in the analysis because of the high degree of uncertainty involved in its initial definition, and in the rapidity of the learning process in space. Turning to (c), it is evident that the costs of the elements required for converting solar power to electricity is closely tied to (b), the manufacturing costs. For the purposes of this paper, only photoelectric solar cells will be considered in order to limit the scope, although the conclusions should be equally applicable to other systems such as closed-cycle thermal engines. The costs of producing solar cells on earth have been following a well developed learning curve, whose projection through the next two decades indicates a potential cost reduction by two orders of magnitude, even in the absence of the manufacturing
RkJQdWJsaXNoZXIy MTU5NjU0Mg==