NASA - National Aeronautics and Space Administration

Home KSC Next Gen Site ___Questions? Comments?

A Review of the NASA Space Shuttle and Human Space Flight Fixed and Variable Space Transportation System Costs

July 14, 2009

Edgar Zapata

Operations Analysis

National Aeronautics and Space Administration, Kennedy Space Center

The following review has been assembled to assist in understanding the topic of the NASA Space Shuttle fixed and variable space transportation systems costs. The data covered can reflect on the range of fixed and variable costs of new human space flight systems derived from Shuttle systems or more generally the existing human space flight systems, processes, infrastructure and contractors.

1.       Space Shuttle Fixed Costs

The Space Shuttles “fixed costs” are an oft quoted number. The topic got increasing attention in the language of fixed, variable, marginal, average and related terminology in the early 1990’s. The quote below from the 1993 General Accounting Office report “The Content and Uses of Shuttle Cost Estimates” places the Shuttles fixed costs at 90% of total yearly costs.

Almost all reference’s since the early 1990’s to the Shuttle’s fixed costs, and likely the one above, have as a source a single study in 1991, the Shuttle Zero Base Cost Study. The entire document is available on the web and totals hundreds of pages. There are key pieces of information in the study that can still be used to reflect on today’s real-year fixed costs of the Space Shuttle. Table 1 that follows is from this study. It reflects “zero base” costs, which are costs relative to the capability to launch just once per year.

Table 1

 “Zero Base” costs and fixed costs should not be confused. The Zero Base value would be a capability to launch once, and any subsequent value assumes that the flight rate capability exists only up to that flight rate, and not above it. By way of example, a Zero Base cost at two flights per year would not include 3 Orbiter Processing Facilities, since only 1 would be required.

Nonetheless the Zero Base study can be used to reflect very well on the Shuttle’s fixed and variable costs by reading the crucial Figure 1 below.

Figure 1

Knowing that history has born out over 100 flights that the truest flight rate capability of the Space Shuttle is close to five flights per year, without stressing the current systems and processes as usual, then Figure 1 would indicate zero base costs at about 80% of total costs. That would be the above intersection of ~$2,500M at 5 flights vs. the ~$2,000M yearly for just the first flight. To further refine towards a fixed and variable concept it can be assumed that the derived variable cost total between five flights and one flight, about $500M (back then, in 1994 dollars), can be divided over the 4 ensuing flights to yield a variable cost of about $125M a flight. Zero flights, usually associated with fixed costs, would then have come out back then to $1,875M a year, with variable costs per flight, up to a 5 flight a year capability, at $125M a flight. Fixed costs would be 1,875 / 2,500 or 75% of total yearly costs.

2.       Updating Previous Study Values to Today

In terms of ratios the major assumptions that would allow using this older data point today, in lieu of repeating the study, would include:

         Assume that evolving changes in the vehicle and infrastructure technology have just replaced past functions in new ways

         Assume that evolving contractual changes have just replaced past functions affecting any totals but not ratios

         Assume the persistence of the functions drives the fixed to variable cost ratios

(These assumptions are tested further ahead.)

For today’s budget the Space Shuttle value and the Space Flight Support (SFS) values can be taken from recent NASA e-budget system data as shown below in Table 2. SFS being accounted for separately is one change since the 1990’s and it can be excluded here as being a general function that Human Space Flight would have to support in any architectural rendition in the future unless it is specifically reinvented.

Table 2

The Space Shuttle’s 2008 budget value in direct program dollar accounting would be $3,310M a year. The value for 2009 is not used as it is already starting to reflect reduced costs as the retirement of the Space Shuttle approaches.

The Space Shuttle’s fixed and variable cost in 2008 dollars, using a $3.3B a year budget would be:

         If 5 flights per year –

         If in FY 2008 dollars -

         Then, total Space Shuttle budget = $3.3B in direct program accounting

         Fixed Cost = 75% = $2.5B/year

         Total Variable Cost = 25% = $825M/year

         Variable Cost Average per Flight = $165M a flight


3.       Testing Assumptions

The assumptions used above can be partly tested to show how an old studies ratio set still applies to the question of the Space Shuttles fixed and variable costs. There are a handful of major changes in accounting, dollars and process from the early 1990’s to today:

         Switchover to “full cost accounting” in the late 1990’s.

                                                   i.      Fewer civil servants booked for as directly charging to the Space Shuttle program, all other’s moved off

         Creation of the United Space Alliance (USA)

         Flight Rate


         Switchover to “direct program dollar” accounting in 2008/2009

Without detailing the mathematics, if an analyst were using the older FY 1994 cost data to predict FY 02 cost data, post USA, post getting into full cost accounting, knowing only 2 things - the drop in flight rate capability, and the drop in civil servants, and adjusting for inflation, the analyst would achieve a result that’s 99% of the actual data (under-estimating). This implies that the savings achieved in the Space Shuttle in the late 1990’s were a result primarily of off-loading 3,700 civil servants (moved to other programs, especially the International Space Station) and of the reduction in steady flight rate to 4 per year (also freeing funds for the ISS). Assuming any other factors, such as efficiency in the consolidation to USA, would throw the estimate off the actual data. This supports the steadiness assumed in the system processes, functions and behavior inherent in the prior assumptions. The switchover to “direct program dollar” accounting is a last adjustment, simply subtracting the percent of Cross Agency Support (CAS) / Center Management and Operations assigned overhead reflected as already removed in Table 2.

4.       Are the Space Shuttle’s Fixed and Variable Costs also the Costs of any Shuttle-Derived System?

An alternate view of the Space Shuttle’s costs is shown below in Table 3. This view is un-official, a merger and filling in of the matrix from assorted sources with assorted modifications and guesstimates. The data for KSC and Florida is the best in absolute values as that has been a focus of this operations analyst.

Table 3

The answer to the question “are the Space Shuttle’s fixed and variable costs also the costs of any Shuttle-derived system?” is more “yes” to the extent more of the existing Shuttle assets above are used, carrying in existing Shuttle processes, flight hardware suppliers and methods, with associated reliabilities, infrastructure and functions.

5.       Closing

The notions of the Space Shuttle’s “fixed costs” can be thought of with some out-of-the-box perspectives that deserve more attention than the time available to address these here:

One-the fixed costs we call Shuttle are really Human Space Flights, a way of doing business, related to ANY Shuttle-like system, supply chain, or human systems, to the degree these are not re-imagined.

Two-fixed costs are mostly a system of processes driven mostly by contractor in-direct functions from requirements generation to requirements fulfillment. Touch labor that touches flight hardware is the smaller part of any contractor workforces. The majority of the contractor resources are in-direct, where the money is. And contractors are the bulk of costs. So consider these fixed costs of processes, not just hard infrastructure ownership.

Third-NASA overhead is no longer in current program accounting. The soup-to-nuts enabling support services in NASA are no longer in direct program dollar budgets. Rather these items such as procurement, finance, information technology, security, general facilities one step removed from flight hardware processing sites, human resources, and safety and mission assurance are under Cross Agency Support (CAS) and Center Management and Operations (CMO) as revealed also in Table 2. So the Shuttles fixed costs are not primarily due to NASA / federal base support, a common misperception.

Fourth-Utilization has been left unaddressed. Fixed costs can be a strength, a valuable asset, ready to produce for only marginally more than is already a given. The behavior of fixed costs, and contractor’s processes and their infrastructure, in this respect must develop some out-of-the-box thinking, such as how to ramp up production of flight elements of sufficient internal commonality to avoid multiplying fixed costs. This would be done while exploiting the benefits to flight rate and exploration of relatively low variable costs. This commonality trade space is usually opposed to optimization of individual flight elements for performance. This need not be the case. Related to commonality is the concept of amortization, the use of existing external capabilities that are not owned nor used exclusively by NASA. The Commercial Orbital Transportation Services (COTS) concept relates to the former as would the use of any Expendable Launch Vehicles. Yet the degree to which optimization and customization result in exclusive use or full ownership of unique assets (flight or ground) will determine the degree to which the quandary of high fixed costs again arises.


Also see:


Website Contact: Edgar Zapata, NASA Kennedy Space Center