Estimating the cost of building and operating a space elevator at this stage of development is challenging, but we feel for at least large segments of the program we can get cost estimates that can guide future decisions. We feel we have good cost estimates for much of the program with the cable production being the largest uncertainty. We have also tried to make conservative cost estimates. The bottom line is that the space elevator could be built at a cost comparable to many U.S. endeavors.
As we have progressed through this study several aspects of our program have changed. The launch vehicle is one of them. Due to results of further investigations into the deployment of a cable from orbit and to eliminate on-orbit splicing of the initial cable we have modified our deployment scenario. In our current plan we will send the entire system to low-Earth-orbit on shuttles, assemble the pieces and then send the large assembled spacecraft to geosynchronous orbit (see Chapter 5: Deployment). The vehicle to reach LEO can be either a standard shuttle (7 launches) or the yet-to-be-constructed shuttle-C (3 launches). To go from LEO to GEO we propose to use a system based on the Centaur for our calculations. The standard shuttle option will cost approximately $500M in current dollars per launch (based on $245M in 1988 dollars) and the Centaurs are roughly $30M each. This gives us a total transportation cost to GEO of $3.7B. If we go with the shuttle-C launch option we will have an additional $3B - $7B for development and $2B less for the four fewer launches. We will get a larger first cable and avoid some of the constraints on launching propulsions systems with the shuttle-C option so it should be considered. The launch costs on the shuttle-C option will be $4.7B - $8.7B. To reduce costs it is also possible to launch some segments on other commercial vehicles. This could save up to $500M in launch costs (an Atlas V/Centaur with 20,000 kg to LEO capacity is $90M) and possibly months on the deployment schedule since the commercial vehicles would use different launch facilities from the shuttle.
The initial spacecraft will be relatively simple in design but one of the larger systems ever launched. The main purpose of the spacecraft is to take its cargo (the cable) to orbit and deploy it. There are no complex optics or electronics, but there are mechanical systems. We believe this spacecraft could be built for approximately $1B.
This is the most difficult component of the program to get an accurate cost estimate for. The cables (208) such as we are discussing have never been produced before. It must be a completely automated process from feeding in the raw materials to final testing. Besides the complexities and quality requirements of the cable itself, large-scale production as we are discussing is nothing new to many industries (textile, automotive, fiber optic, electronic, etc.).
The 207 climbers we intend to send up the cable will all be the same basic design but need to deal with the ever increasing cable sizes. Hopefully the climbers can be designed in a modular form so that for some large number of climbers the same components can be used just in larger quantity. For example, the first 10 climbers may have five motors to pull them up the cable. The next 10 may have several structural components and one additional motor added without substantial modification to the rest of the climber. If designed properly each of the 207 climbers will not be a custom production. Based on small spacecraft work we believe that $20M per climber (excluding the cable cost) is a reasonable estimate with the first climber being more expensive and the later ones being less expensive.
Radar systems such as the Haystack Observatory have begun tracking objects down to 1 cm in size. Optical systems, which have some advantages, (and disadvantages) are also coming on line. NASA has already requested that the current space surveillance network be upgraded to track debris down to 1 cm with improved accuracy. It has also been suggested the upgrade include moving the facilities to the equator. More details are required, but if the upgrade were to occur, it may satisfy all of the foreseeable tracking needs of the space elevator. If the upgraded space surveillance system does not come online before a space elevator is undertaken then a set of tracking facilities (radar similar to Haystack or a phased array or optical as suggested by Ho39) would be required to be built. A reasonable estimate is that five new facilities would be required along the equator. Five facilities would allow good tracking and not place the elevator in jeopardy if one or two tracking facilities are down. Based on estimates for a new radar observatory each facility could cost less than $100M (Berkeley's One Hectare Telescope for example is $25M) including high-speed computing facilities.
Johnson Space Center has also conducted a study on constructing a new set of tracking facilities40. The system that was proposed used current technology (X-band phased arrays and dishes) and U.S. locations. The total estimated cost of one billion dollars to build and 100 million dollars per year to operate were given without any breakdown but are roughly close to our estimate above.
We will use what we feel is a conservative cost estimate of $1B for construction. Operations will be broken out with the rest of the program belows.
If a platform modeled after the Sea Launch facilities is selected, the total cost of this system would be less than $300M. This does not include a power beaming station (see below) or unique new facilities that may be desired such as a nearby, floating airstrip.
The current engines on-board the Sea Launch platform generate 20 MW of mechanical power. A similar system could be used to generate the electrical power needed by the laser beaming system. If the total efficiency of the beaming system is greater than 5% (see power beaming section) then this system could supply 50% of the power required for operating a 20 ton capacity elevator. Either a larger power generation system or two will be required for our purpose but for now we will use this system for our baseline. We will need a separate beaming facility in addition to the one at the anchor and will probably want a backup system at each facility. For a budget estimate we will use $100M (a very conservative estimate based on the Sealaunch program) for each of the four power generation systems. The total power generation budget estimate is $400M.
For our cost estimates we will use a laser beaming system. The beaming system will consist of the facility infrastructure, high-power lasers and a large deformable mirror. We will want both the power beaming facility located at the anchor and a separate beaming facility on a second platform on land to improve the transmission duty cycle. The facility costs for the separate beaming station will be the cost of the platform ($300M) if an ocean station is selected and possibly $10M per land site. The facility costs at the anchor platform could be an additional $100M for modifications. The Compower power beaming system is well along in its development and has accurate cost estimates for that system20. The lasers are $100M each (fixed price from Berkeley for a 1 MW system) and the deformable mirrors are $125M each including infrastructure. We will need 4 MW of output laser power for the initial cable so each beaming station can be assumed to have 4 laser systems and between one and four mirrors depending on how well the beam paths can be combined. If we assume the beam paths can not be combined we will need four completely independent systems. Each laser beaming station will then be $900M ($525M for single mirror system) if no savings are assumed because all the systems are identical. This would mean we should conservatively budget $2.2B ($1.45B for single mirror systems) for the two beaming stations with each on an ocean platform.Operations
At this point, we will include operations for the first ten years of the space elevator. It is obvious the cable operations will go well beyond this but this will cover everything we have discussed for the initial cable. JSC's estimate for operating the tracking facility is $100M per year31. There was no breakdown on this cost, so we will take it at face value. For operating the space elevator we will assume two platform facilities with crews (20 per facility), technical staff (30 per facility), support personal (30 per facility), and administration (10 per facility). This is a total of 140 people (this is roughly equivalent to the Sealaunch program). If we take an average salary of $100k and 100% overhead we have a total yearly operations budget of $28M. We will also assume there will be personal located back in the states of equal numbers. This gives us a yearly personnel budget of $56M plus the $100M slated by JSC for the tracking facilities.
At this stage of a program accurate cost estimates are difficult especially for something as unique as the space elevator which really has no prior completed project for comparison. We have implemented a 100% cost contingency to cover items that have been overlooked or may show-up later in the program and to cover growth in costs.
The summary of our budget estimate is in table 11. 1. This study has clarified the cost of many of the space elevator subsystems (relative to the original estimates in our proposal). There is still uncertainty in the cost of the cable production, on-orbit operations, and some in the initial spacecraft and climbers. Due to the uncertainties in these areas we have increased our contingency. These estimates are based on technical requirements, political costs are not included.
|Original Component||Original Cost Estimate||Revised Cost Estimate||Comment|
|Launch costs to GEO||$1.4B||$3.7B||Changed to 7 shuttles with Centaurs|
|Cable Production||$5B||$5B||Still difficult to estimate|
|Spacecraft||$1.2B||$1B||One large SC instead of 4 small ones|
|Climbers||$1.25B||$4.2B||$20M instead of $5M each, 207 instead of 250|
|Power beaming station||$10B||$2.2B||Based on Compwer system, two systems on ocean going platforms|
|Power gen. station||$4B||$400M||Based on Sea Launch program's power generation system|
|Anchor station||$5B||$300M||Based on Sea Launch program's ocean going platform|
|Tracking facility||$1B||$1B||New facilities based on several existing programs and a NASA study|
|10-year operation||$1.56B||Now broken out separately|
|Misc. and cont.||$10B||$20B||100% contingency|
Several additions can be made to the overall space elevator program that would be of considerable benefit which include:
|Component||First Cable||Second Cable|
|Launch cost to GEO||$3.7B||0|
|Power beaming station||$2.2B||$1.6B|
|Power gen. station||$400M||$400M|
|Misc. and contingency||$20B||$5B|
After the first cable is complete, the second and following cables will be considerably easier, faster and cheaper to build per kilogram of capacity. Climbers can be sent up the first cable deploying and building the second cable as they climb. In this case a cable of comparable size to the original can be made in 7 months instead of the 26 months of the first. If we assume we work out some of the designs and difficulties in building the first cable and eliminate the non-recurring costs we find the second cable can be built for under $15B (table 11.2). The third and subsequent cables will cost less than the second. The value of these cables are no less however and, if it were so decided, some of these subsequent cables could be sold as completed units to private industry or other entities to recoup the cost of the entire space elevator program. When considering larger cables the scaling of the climbers and power beaming systems must be taken into account.
|ç Chapter 10: Challenges||Table of Contents||Chapter 12: Schedule è|