designs are so unreliable (and systematically low) as to be of little value except for relative evaluation of options all estimated with the same ground rules and methods. The "Back Door" Approach There is a “back-door” parametric cost analysis strategy that offers much insight. This strategy [1] infers from the parametric estimates, the labor content of SPS hardware manufacture; [2] assumes that the labor content does not change from Earth to space; [3] calculates from the reduced transportation cost for space-derived materials, the break-even delta cost for support of space workers. In other words, if transportation cost is reduced, one can afford a higher cost for production labor; how much higher? This technique will overestimate space manufacturing costs: it implicitly assumes that the same level of automation is justified for space and Earth manufacturing, whereas the higher labor cost in space will increase the economically optimal level of automation. A word of caution: it is not plausible to assume that automation can bring space manufacturing costs below terrestrial costs. If we assume this is true, then one can take the space manufacturing level of automation (it will cost less on Earth) with Earth-based labor costs to arrive at an Earth manufacturing cost less than the original Earth-based figure. Hence, the original was not optimal, whereas it is unlikely that a level of Earth-based automation far from optimal would be adopted. Table 4 presents a summary of the SPS hardware as defined by the systems studies (11). Table 5 presents a summary of “break-even” labor costs, ranked in order of delta cost and presents the SPS hardware items in a different order than Table 4. This is only the result of a sorting procedure; the data were not altered. Figure 5 plots the data from Table 5, showing the economically-justified lunar materials (space manufacturing) labor level as a function of delta costs to support spacebased labor. One would expect a discrepancy in labor estimates between the space manufacturing studies and the Earth-based reference, but the figures represented in Table 5 (total labor of about 80,000 per 5-GW SPS per year, or about 170,000 for the reference scenario) approach two orders of magnitude more than the space manufacturing studies. One would not expect a discrepancy this great. Discussion and Analysis A significant uncertainty, of course, exists in the reference system inferred labor content. Some additional explanation of the technique and a few.examples will add perspective: The Boeing Parametric Cost Model (PCM) estimates costs beginning with the major component level (level 6 of the WBS) and builds upward to obtain the total program cost. Cost estimates are based on physical and performance parameters at the hardware level and programmatic parameters (quantities, learning curves, production rates, etc.) at the project level. This methodology thus mirrors the actual approach used to develop and produce aerospace hardware. Boeing historical data collected in the Estimating Information System (EIS) data bank provide the raw information from which functional man-hour estimating relationships (MER) are formed. The procedure used to establish production costs is implemented in the same PCM program. The actual production cost of the first few units is very difficult to account, since
RkJQdWJsaXNoZXIy MTU5NjU0Mg==