guest blogger john hare
Recent events make it easier to describe one scenario for getting orbital costs down. This specific example almost certainly won’t happen and just stands in for the dozens of possible ways that orbit could become affordable.
Masten and XCOR have a joint venture for developing a methane lander. Consider the possibility of future cooperation. If in five years Masten and XCOR are operating profitable sub orbitals out of Mojave, it will be past time to start on orbital hardware. If both companies have 300+ kg payload capacities to above 100 km, then the Lynx will have the capacity to lift one of the earlier Masten vehicles (Xombe 0.5?)to above 100 km. A fairly rapid partially retrograde development could have the small Masten vehicle capable of very long suborbital flights if properly assisted. Say, Mojave to Kodiak, Oklahoma, or New Mexico. The Lynx lofts the Masten vehicle to mach 3-4 at high altitude and stages. The Masten test article finishes the boost for one of those spaceports while the Lynx returns to it’s runway.
The experience gained in test flights with almost existing equipment will go a long way toward convincing investors that orbit is well withing the reach of a conservative three stage to orbit transport. (Pete suggested this as two stage over on Transterrestrial) The reentry problems can be explored for mach 10 to 15 including turnaround time and environmental impact. Reentry noise can be one of the largest hurdles for inland recovery of second stages. If, and only if, inland recovery can be done in an acceptable manner will this concept work. A tiny smallsat stage might be launched from the Xombe 0.5 if required to make the point.
For a second set of tests, Masten launches a three barrel vehicle with their full size suborbital craft. The two outers deliver the same assistance to the center vehicle as Lynx does to the smaller craft. The full size suborbital vehicle lands downrange at one of the available spaceports. Once this is explored, a 300 kg upper stage is used to place a useful smallsat in orbit. Some testing is done on recovering the upper stage.
If the above tests confirm that the technical concept is sound and acceptable to the uninvolved public at the recovery spaceport, then funding can be sought for a full orbital system. A much larger XCOR vehicle (Panther for this post) is developed to launch the full size three barrel Masten (Xorbit for this post) assembly. This time the two outer vehicles cross feed propellant to the center vehicle and land at one of the downrange spaceports. The center vehicle goes on to deliver 300 kg of payload to orbit. This allows XCOR to focus on the much larger suborbital Panther while Masten focuses on improving existing vehicle performance and reentry. The Panther could also be the rent-a-booster as Clark suggests over on Space Transport News. The technical risk to orbit from that point would be comparable to funding XCOR and Masten to a full suborbital operation now.
A launch organization has four main costs. Development costs. New vehicle costs. Fixed operating costs. And marginal costs. Only marginal costs are mostly unaffected by flight rate. On a gas-n-go vehicle propellant is a major marginal cost.
On this three stage to orbit with each stage having a mass ratio of three, (total MR=27) there would be about 36,000 kg of propellant or about $18,000.00 per launch. Other marginal consumables would double this to $36,000.00 per launch or about $120.00 per kg in marginal costs.
Starting from a base of operating vehicles and recent development experience with an intact team, I speculate that the development cost would be about $250M. Expected interest on the money for this high risk field could easily be in the 25% range for a development cost of $65M per year to service the debt. If there is one flight every two weeks, $2.5M per flight to service the debt. At two flights per week, this drops to $625K. With multiple airframes flying twenty payloads per week, this drops to $63K per flight to service the debt.
New vehicle costs for the four airframes (Panther and 3 Xorbits) could be in the $20M range after the bugs are out. At a flight every two weeks, it would take $770K to pay off a set in a year. At two flights per week per airframe, it drops to $193K to pay off in a year. If some financing can be arranged, then costs can drop considerably more. (It could be much less in the reality as the Panther could lift 20+ times per week while the Xorbits could possibly do two each.)
My first guess on fixed operating costs is $30M per year keeping two or more spaceports fully operating with all support staff, including transporting the Xorbit stages back to Mojave. These numbers are based on this guess. At a flight every two weeks, $1.154M per flight. At two per week, $289K per flight. At twenty flights per week, $58K per flight figuring that staff doubles at that flight rate.
flight frequency debt vehicles fixed marginal total
two weeks $2.5M $770K $1.154M $36K $4.46M
two per week $625K $193K $289K $36K $1.143M
twenty per week$63K $193K $58K $36K $350K
If it worked out like this, then a flight rate of once every two weeks would break even at a little under $7K per pound, while twenty flights per week could get it down to around $530 per pound. A lower flight rate under every two weeks would make the plan unworkable while a higher flight rate than twenty per week would make it somewhat cheaper than $500.00 per pound. These numbers assume that interest only is paid on the development and that the vehicles must pay for themselves in one year.
This is very much a first generation space transport. Obvious places to get the cost down further are to reduce development costs while still getting an acceptable vehicle group. Getting a better interest rate. Improving propulsion and dry mass fraction to get the mass ratio down from 27 to 16 or so with less dry mass to lift in the first place. Getting it down to two stages to orbit to eliminate supporting downrange spaceports and the extra stages. Flying each airframe more than the twice a week I have here. Reducing the cost of new airframes and spreading the payback time over more than one year. Reducing the number of support personnel required per flight. And so on to get costs insanely low compared to the present day.
Developing orbit into an economic driver is going to take a lot of work and intelligent handling. Orbital development to date with existing transportation is a mere shadow of the possibilities given transportation that is economical, reliable, and convenient. Everything changes when a decision can be made and executed reliably and affordable in a matter of weeks or possibly days. When was the last time you scheduled anything years in advance? How many times in your whole life do you schedule something years in advance and know that it might cato before it gets started? That is what the whole launch industry is forced to do at this time.
Basing on this cooperative venture creating a somewhat convenient transportation method to orbit, speculation on markets becomes possible. The first markets are somewhat limited in orbital inclination by the restrictions of the downrange spaceport requirement. The choices are a somewhat retrograde high inclination orbit, or a normal orbit with an inclination at Mojave’s latitude. Anything to ISS is out due to inclination restrictions. Tourism is out due to the small vehicle size. What can you do?
While proving the vehicles, expensive payloads are out, so the focus must be on things that have low intrinsic value with only being in orbit making them valuable. Propellant is the first obvious choice. The first launch to each available inclination carries the largest propellant tank that it can as payload with subsequent launches filling it up. Or possibly one of the ELV companies leaves an upper stage in the right orbit that can serve as a depot. This is only good if there is a market for propellant from that particular orbit though. I dislike the concept of buying propellant, or water, or sand in orbit for the purpose of artificially creating a market.
The first stopgap depot would need to be in a normal orbit at Mojave’s latitude. When it has sufficient propellant delivered to make it worthwhile, one of the majors launches an un fueled GEO bird from the Cape to rendezvous with it. A 25 ton GEO bird unfueled could take on 50 tons of propellant before boosting to it’s final orbit. This would make for a very massive satellite compared with the current operational GEO sats. This would take about 170 launches of Panther/Xorbit. A contract for a million a launch would be profitable to both sides. The GEO provider would have the capability of a 75 ton launch vehicle for an extra $170M per GEO sat while only risking a 25 ton launcher. The Panther/Xorbit companies would need to get the launch rate above 3 per week to make a profit and pay off their debt. The 3 launches per week to reach profitability would take over a year to complete for each GEO bird that takes advantage of the capability. Assuming there are at least two of these monster satellites per year to refuel, the launch rate is at nearly 7 per week. If deliveries could reach that flight rate, about 40% of each launch would be for profit or paying off the debt. Debt payoff would be under 3 years and would further reduce the launch costs by a factor of almost two, which would make the launches even more profitable at that price point.
Propellant is not the only possible cheap on the ground expensive in orbit payload though. Military surveillance would benefit from the ability to place fleets of cheap observation satellites in orbit. The gaps in coverage of the existing sats are most likely known to the people with something to hide. They probably time most of their sensitive moves to take place in these gaps. The military could easily spec an LEO surveillance sat that could be mass produced for under half a million. Though these birds would be far less capable than the high end machines they use now, a thousand of them would provide wall to wall coverage. At a million a launch and half that for the actual satellite, they would be looking at $1.5B for the whole program not including collecting and integrating the data. Allowing three years for placing the constellation, flight rates and costs match that of the propellant market only. Under three years to payout and higher profit thereafter.
Communications is frequently brought up as the target for early market. With cheap reliable launch as suggested here, less capable LEO satellites could be used in constellations with spares on orbit and on the ground ready to be launched in days if required. It is quite conceivable that these 300 kg comsats could be mass produced for a half million just as the surveillance satellites are. A billion and a half for a robust LEO comsat constellation of a thousand birds should be a business plan that would close. A three year launch schedule has the same numbers as the propellant or surveillance constellations. Under three year payoff and 40% profit after at a million a launch.
Astronomy and microgravity science and Earth observation satellites should be a market to match the last two. Though the individual capabilities are less with the smaller satellites and cheaper construction, a cast of thousands could make up the difference. The ability to launch follow up or replacement experiments in a week or less would make commercial use of LEO much more user friendly. Over a hundred countries with thousands of universities and tens of thousands of private companies would be the market this time. With similar launch rates to the other users, the money works out the same.
If all four of these markets were to emerge though, a flight rate of 30 or so per week would drop costs still further. Costs of under $350K per flight would allow a charge of a half million per flight to make over 30% profit available for the debt payoff or investor ROI. It would also create even more incentive for the users of the service to make less expensive satellites for an even lower total cost. $4.5M per week profit would pay off the development debt in just over a year.
The assumption that these satellites would be inferior to the $Dirksen Galactica models launched by the current providers might not be true. The current birds are just too expensive and hard to replace to fail so every possible effort is spent making them reliable. This means literally gold plating some components and using only hardware that has been “space rated”. The time required to space rate components added to the long lead time on the current launch capability means that by the time any component reaches orbit, it is incredibly expensive, and it is obsolete by standards on the ground. The point has been made many times by many people that many of the current satellites, especially comsats, are so expensive that the cost of launch is not all that important. With replacement launch on demand, satellite reliability becomes somewhat less important. It becomes possible to launch the latest electronic devices without “space rating” them at all. If they survive, they are space rated, if not, they are not. Henry Spenser has noted that the commercial stuff launched in the Canadian satellite that he worked with has lasted for years. How much capability is in $10K worth of May 2010 commercial electronic components components compared to $100M of “space rated” 1995 electronic components?
When failure is not an option, success can get expensive. We have all read that. How many have also noted that when failure is not an option, success also gets much less capable?
With launches available on demand for half a million, many high risk or complicated capabilities can be tested and possibly implemented. A 300 kg tug could do considerable work in LEO. It is past time to deploy tethers for extensive testing. Debris cleanup vehicles can be tested and then employed. Fairly small service vehicles could do good work in GEO. National Geographic could afford a 300 kg Lunar, Mars, or NEO probe that was launched on one flight and fueled up by others.
It just gets better and the sky is not the limit.