As far as implementing this idea, the technology isn’t the hard part. Technologically, this is something that could’ve been done in the 70s. Modern technology and modern launch services make it a whole lot easier and more feasible, but the technology isn’t the key obstacle. Money is and always has been the biggest obstacle. But I think I have an idea, and it’s just crazy enough that I want to share it.

Any business plan whose first step is “first we convince a billionaire to give us lots of money” usually deserves to be laughed off the stage. But this isn’t a business plan competition entry, or some pitch before VCs that I’m demanding to be taken seriously, so I’m going to suggest just that. Even with a wealthy philantrocapitalist, I think you’d still want a concept that both gives you a reasonable chance of making the money back if things go well as well as minimizing your losses if it doesn’t work out.

Anyhow, this is a bit of a long-shot, and definitely not fully-baked, but here’s what I have so far. The business case revolves around a few core concepts:

• A privately developed simple lander and an ITAR approved method for launching it on both US and domestic launchers.
• Using barter with various space agencies with domestic medium-lift vehicles to provide both the startup launches and the sustaining launches
• Making revenue off of selling remaining space to corporations, research institutions, and smaller countries that are interested in lunar experiments, but lack indigenous launch capabilities
• Possibly offsetting initial lander development by selling rover delivery services to NASA or other large space agencies.

Some of these sound a bit crazy, so why don’t I explain them in turn.

Private Landers
The key technology piece in the project is obviously the lander. As discussed before, I’m thinking of something in the 10-20klb IMLEO range, with a payload in the 4-6klb range. The propellant combination for the lander doesn’t hugely matter. It could use storables like Martijn likes, it could use space storables like LOX/Methane or LOX/Propane. Heck, it could even use LOX/LH2. While the state of the VTVL industry isn’t quite mature enough where you could just order one of these custom and have it delivered to your launch pad 6 months ARO, a lander in these capability ranges isn’t a huge stretch for the commercial space industry, especially if they can partner wisely with some of the more traditional space companies or work with NASA via Space Acts. DC-X was actually a much bigger, probably more complicated system, and was done by a traditional aerospace company for around $100M in current dollars. A bare-bones lander, developed leveraging the emerging capabilities in the entrepreneurial community could probably be fielded for less than that. Possibly in the $50M range. You don’t need to push too hard on mass fractions or engine performance (you need to push a bit, but it isn’t as weight critical as some of the Apollo LM systems), and the technology is a lot more mature than it was in the 60s.

An important part of this process is not just developing the lander, but also working from the start with ITAR to make sure a process is in place that will allow you to launch on as many international launch vehicles as is feasible. This may not be fun, but is probably doable with appropriate precautions.

International Horse-Trading
Most space agencies prefer to spend money within their own borders, and interact with other agencies on a barter basis as much as possible. While this can sometimes lead to suboptimal solutions, it might just work in this situation. On the launch side, the barter would go something like this–the private entity would provide a lander, all lander ops, and physical launch integration work, and the space agency (NASA, ESA, RSA, JAXA, ISRO, or CNSA) would provide the lifter and upper stage for the mission. The launching country would get a certain share of the lander’s cargo space for their own experiments, a certain portion would be reserved for consumables and spare parts, and the remainder would be owned by the private entity to resell to other countries without launch capabilities (say a 40/40/20 split). In addition to transportation of the space hardware, the launching country would also get a share of the astronaut’s time on the surface. So basically you’re providing them with transportation and manned experimentation on the lunar surface in exchange for them providing a launch done by their own people. If one of the countries is willing to take some additional risks, they could even “buy” one of the two initial astronaut slots, in exchange say for a commitment to a certain higher share of the logistics launches per year. In exchange they’d get both the prestige of having one of the initial lunar crew, as well as a higher share in the available time. Over time, as the risk decreases, the initial crew could also be expanded (once again on barter terms that would have the agency in question shouldering a larger share of the required launches).

It should be mentioned how crazy of a bargain this really is for them in comparison to the typical lunar mission approach. Look at Constellation. It will be a lot more capable, but ultimately, somewhere around $10B/yr (and about $150B up-front), you get 4-person years/yr (2x 4-man crew rotations) and about 75klb of cargo (2x 17mT landings) on the moon once you have a base setup. Calling it a 60/40 split on costs (for manned vs cargo flights), that comes out to $1.5B per person-year, and about $53k/lb on the lunar surface–ignoring development costs. With a program like this, say you gave a country 1/4 of a man-year per launch, and about 1800lb, at a cost to them of call it a $200M launcher plus extra upper stage for the transfer. Splitting that $200M the same way (60/40), that gives you $480M per person year, and about $45k/lb on the surface. You don’t save a huge amount per pound of cargo on the surface, but your cost per person hour is about 1/4 as much (which is once again not too surprising–you’re not rotating crews, and not having to carry enough propellant to get them home–which takes about 4x as much mass per mission compared to a one-way manned landing). And you don’t have to spend tens of billions up-front, and you can buy your lunar program “by-the-slice”. Paying for an extra launch every year (and some lunar systems costs) is well within the budget capabilities of many of these agencies. While they might not be willing to take the risk of flying their own astronauts, or of “owning” the program, they are a lot more likely to be interested in a program like this, where someone else is shouldering the key risks, and they’re just getting a cheap deal. Even if they have their own lunar ambitions down the road, using a service like this would allow them to drastically reduce their technological risk moving forward, and might allow them to get a lot more benefit out of their investment when they eventually get that capability themselves.

“Sovereign Customers”
One of the key markets Bigelow is looking at for his inflatable space habitats is providing smaller countries with a way to participate in space for much cheaper than trying to do everything in-house themselves. By lowering the cost to participate, it makes it a lot more feasible for smaller countries, and even some corporations or research institutions to participate. This may be a country like South Korea wanting to send a rover that can get maintained by the astronauts over time. It may be a country wanting to do its own sample return mission–with the ability to have a human on the ground helping to presort/preprocess samples to maximize the bang for the buck. It could be a company like Catepillar that wants to get involved in lunar surface systems for future exploration programs sending a bunch of bearing concepts to test exposed to the lunar environment. It could be some small startup that has a crazy idea for lunar dust mitigation that it wants to try selling to future government programs, but needs testing and debugging first. There are many possibilities. The key here is that since the launch is already paid for, the private entity running all this can price the payloads however makes the most sense. You do need to cover lander costs, ground-ops costs, and the time of the scientists, but it might be possible to offer these slots at a price that is lower than they could buy commercially to try and stimulate demand, or if there is enough demand already you could price it high enough to make a decent profit. If there’s enough demand, you might even be able to justify paying for an additional “purely commercial” flight or two per year. You would want to save up some of the money to cover contingencies–like if something breaks down and you have to fly an emergency resupply flight on short notice, or if you decide for one reason or another to throw-in-the-towel after a few years, you can send enough propellant to get the settlers home. But depending on the interest level, this could easily be a business that has revenues in the low hundreds of millions per year.

Minimizing the Initial Risk
One additional market for the lander, and one that could allow the initial investment to be recovered a lot faster, would be to see if you could sell it to one of the space agencies for landing a rover or some other scientific package. The key here is that the lander is getting developed, on the philantrocapitalist’s own dime regardless of if he can presell any lander slots. This makes it easier to sell it as a commercially available service instead of a government funded development program. Using a light Atlas vehicle for instance (maybe with one or two strapons) you could probably short-load the vehicle enough to put a couple hundred pounds of useable payload onto the lunar surface. For a bundled price of say $200-250M for the launcher and lander, it would still be a steal transportation-wise for your customer, but could possibly pay off the initial costs of the project in one shot, even before the initial landings. The good news is that while its great if you can presell the landers for other applications, it isn’t the end of the world if you can’t.

One other way of minimizing the downside may be to see if you can prearrange the initial several launches. If you can line up enough international partners, it may be possible to get the initial setup done without having to actually buy any of the launches yourself. You’d still have to pay for the landers, but this way your total capital at risk for the startup is only the cost of 3-4 landers.

Anyhow, comments? thoughts? attempts to send nice young men in their clean white jackets to cart a certain space blogger away?

Let’s turn to the Augustine Report itself for some information. On pages 83 and 84 they discuss implementing the Program of Record on entirely unconstrained budgets–ie if we gave the program the full funding it needs to execute, and allot it to move at the full pace it can realistically move at, what do we get?

• A $145B pricetag over the 2010-2020 timeframe, which doesn’t even get us to the point of having Ares V and the LSAM ready for operations, much less a moonbase.  This would require almost $5B extra per year–ie a 25% increase in NASA’s topline budget.
• An international space station deorbited within 5 years of its completion, during which time the only method of access would be by paying the Russian government for flights.
• A crew launch vehicle that becomes available two years after its first destination is deorbited, and whose operational costs have to be carried for over half a decade until we have any of the tools that would be necessary to actually use it for anything.  But don’t worry, we can spend $2B+ per year to send even fewer astronauts flying in even more useless circles.
• A seven plus year manned orbital spaceflight gap in the US.
• Almost no investment in long-term technology development (not much more than the current SBIR budget, and entirely focused on short-term Constellation needs, not on making future missions safer, more affordable, and more valuable).
• No stimulation of commercial industry beyond the CRS contracts which wouldn’t be extended since the ISS would be gone by 2016.  No investment or early market for commercial crew delivery
• No money to actually develop hardware for actually doing anything on the Moon, since almost all of the money will go to figuring out how to go there while maximizing employment in Shelbyville.
• No more robotic orbiters or landers for years to follow-up on the work LCROSS did.

But hey, at least if we do it this way, sometime 15+ years from now, we’ll have the ability to send 8 people to the moon every year at the cost of an “exploration” program that costs almost as much per year as NASA’s entire current budget!

If you assume that there are parts of NASA outside of Huntsville that actually matter (ie that NASA != Northern Alabama Space Administration), the situation gets even worse.  In order to fund Constellation at full speed without splashing the space station almost as soon as it’s completed, you would need $159B over that timeframe, which constitutes a $7B per year increase for NASA.  That increase still:

• Gets you a space station you can’t access without the Russians for most of its operational lifetime (why does Congress trust Russian commercial space more than American commercial space, btw?).
• Gets you no real investment in long-term technologies, ensuring that the cost, safety, and efficiency of manned spaceflight will be stagnant for another couple decades.
• Gets you no real investment or encouragement of the commercial industry (in direct contravention of the laws of the land and NASA’s charter I might mention).
• Gets you no more robotic follow-ons for LRO and LCROSS for over 15 years.

Compare this with the Flexible Path option that Mark likes to mock so much.  For less than half as much of an increase per year, you get:

• Robust ISS utilization through 2020, with multiple methods of providing crew and cargo delivery that aren’t all dependent on Russia
• Investments in commercial space that can help keep the US in the forefront of space technology and utilization
• Robust investments in high-payoff medium-term technologies like propellant depots, space radiation, space nuclear power, aerocapture and other EDL techniques, ISRU, and other high-payoff technologies that can vastly lower the cost of future exploration missions, allowing us to accomplish more for less and at lower risk.
• A manned lunar landing program that at most is only 3-4 years behind the current PoR, but when it gets there, it provides a much more affordable, more commercially and internationally interesting program, and has much greater capabilities once you get there.
• A manned spaceflight program that is much more capable of exploring the whole inner solar system, and not just doing a few flags and footprints landing on the Moon.
• A manned spaceflight program that builds on and leverages our impressive achievements in robotic space exploration.
• A program that in spite of doing a lot more looking, also allows a lot more touching of new destinations like NEOs and Phobos/Deimos, all on about the same timeframe that the PoR would at best be going for its first lunar landings.

Where I come from, we tend to think that getting a heck of a lot less while paying a heck of a lot more is usually the sign of a sucker.  I just wish that a few space pundits and public figures didn’t keep enabling Senator Shelby and his ilk from hijacking NASA’s budget to enrich his campaign contributors at the rest of our expense.

[Note: As an aside, am I the only one who finds Shelby's latest childish tantrum accusing the Augustine Committee of being compromised by biased by evil commercial lobbyists to be richly and hilariously ironic?  When it comes to lecturing people about the evils of lobbyists corrupting the political process for their own personal gain, Senator Shelby has about as much moral standing as Tiger Woods does when it comes to lecturing people about marital fidelity.]

ISS Centrifuge Accomodations Module (Credit: NASA and Wikipedia)

As I’ve discussed before on this blog, our knowledge of the impact of gravity levels other than microgravity and 1 gee are almost virtually nonexistant.  We have billions of data points at 1 gee, and we have hundreds of data points in microgravity, but we have a few tantalizing hints from the six Apollo lunar landings–nowhere near enough data to make responsible projections.  It may turn out that only a little bit of gravity can go a long way (if the negative effects are driven by fluid distribution in the body like I think it is), or it could turn out that even Martian gravity isn’t enough.

This is the kind of information we really need to learn if we’re ever going to be a spacefaring society, and CAM would have provided that data.  Unfortunately, in 2005, the partially completed module was canceled, due to budget overruns and issues with trying to schedule a launch before the Shuttle was to be retired.  The module has been sitting outside at a space center in Japan ever since.

At the time, that may have sounded like a reasonable decision, but now that it is looking like ISS won’t be splashed in 2015/2016, and with the new emphasis on manned deep space exploration, it would be nice if that decision could be undone.  The Augustine Committee mentioned research on the impacts and mitigation of reduced gravity effects on the human body as one of the reasons for extending the ISS’s operations to 2020.

Unfortunately at this point the team has been disbanded for long enough, and the hardware exposed to the elements long enough that resuscitating the CAM is probably not in the cards.  More importantly, like most other ISS modules, CAM was designed to be launched on the Space Shuttle.  While it is possible to develop an adapter for EELVs that could allow ISS modules to be launched on them, such a system has yet to be funded.  So for now, it looks like restarting the CAM project as originally formulated is probably a dead end.

So here’s my crazy idea.  What about modifying a Dragon capsule to house the centrifuge experiment and its supporting equipment racks?

Dragon Berthed at ISS (Courtesy NASA and SpaceX)

Most of the volume in the CAM design was actually storage rack–10 of the 14 ISPRs in the module were set aside for storage, and only 4 were planned for science.   The actual centrifuge itself was about 2.5m in diameter, but not very thick.  Looking at the Dragonlab datasheet, I wonder if it would be possible to make another copy of the centrifuge assembly itself and fly it as a payload on a DragonLab flight to ISS.  Looking at the available volume, it looks like you could fit the centrifuge and possibly as many as 1-3 of the 4 payload racks that were originally slated for the science mission.  I’d need to do a quick CAD model to see if the ISPRs would fit as-is, or if you’d need to go with some other science rack configuration.  And such a setup wouldn’t have all the capabilities that the original CAM had, but it would give you some of the most important functionality.  Also, being part of a reenterable spacecraft, there would be the benefit that you could bring the setup back to earth to repair, modify, and upgrade it from time to time.

Looking at the sizing, this might require a dedicated Dragon airframe for the project.  The centrifuge assembly itself is too big to fit in the door in one piece!  It might actually be necessary to build the capsule around the centrifuge.  But there’s enough experimentation that would need to be done over the years, that it would probably make sense to do it that way.  The duration of DragonLab missions are listed as up to 2 years.  At that rate, you could do long-duration experiments, but still have the thing back down for maintenance, upgrades, and refitting frequently enough that it might allow you to make some design simplifications.  Also, basing the CAM inside a Dragon capsule would mean that the team designing the science hardware could focus just on the experimental apparatus, instead of having to design a full spacecraft like the original CAM.  That might save a lot of time and money compared to trying to complete the original CAM. Lastly, the Dragon capsule can be either docked to the station, or can serve as a freefloater, whichever makes more sense scientifically.

[Note: It might also, just barely fit with the Orbital Sciences Cygnus spacecraft.  I don't have internal dimensions for that, but judging from the external dimensions, there's a chance.  If someone who has data on what the layout of the usable volume for Cygnus is, that would be helpful.]

By offloading all of the work other than the apparatus itself, and by using a relatively inexpensive launcher, this could be a way for international partners to contribute.  Either Japan could provide the apparatus, or if they’re not interested, this could be a way to involve India or China in the ISS program.  It would also be something well within the capabilities of Canada or the UK as well.

In fact, it might even be possible to do this entirely as a commercial venture or as a privately funded not-for-profit venture.  A commercial venture would be risky, since you would need some sort of guarantee that someone would actually pay for the data.    A not-for-profit with a wealthy benefactor who would be willing to subsidize the experiment, (much as has been done for many university science labs, telescopes, and other not-for-profit scientific facilities) might make more sense.  Imagine the Stanford or Caltech or MIT Orbital Centrifuge Lab, or the Bill and Melinda Gates Orbital Centrifuge…

As far as a spacefaring society is concerned, CAM would’ve been one of the most useful experimental hardware on the ISS.  It may be too late to restart the CAM module as originally conceived, but Dragon–if successful–may provide another chance at making the ISS truly relevant.

Anyhow, Jeff managed in 25 minutes to address human rating, depots, whether or not we need heavy lift, technology maturation and R&T investment, and the need for NASA to find new ways to interact with business. I don’t think he could have hit more of my hot-button issues in 25 minutes if he had tried.

Anyhow, I hear that the HSF committee will have video of today’s proceedings up online soon (possibly tonight) for those who didn’t get up at 5:30am PDT to watch. I’ll comment more later.

Whew! I haven’t had this much hope for this nation’s space program in years!

[Update:  Here's the link to the subgroup's presentation (warning, it's a 14MB powerpoint presentation).  All of it is interesting, but Jeff's part starts on page 76 and goes through page 89.  Chris Chyba's section was the last three pages.]

Before I get into my rebuttal of Rob’s post though, I had an interesting side note worth mentioning.  In thinking of what to write, I decided to start by trying to figure out where I first became interested in propellant transfer and propellant depots.  I’m pretty sure it was after the Return to the Moon conference in 2004.  Because at that time I was still looking at “tinker-toys” style multiple propulsion module approaches to lunar transfer systems.  So, I did a search on this blog for propellant transfer, and realized that all the way from my very first substantiative post on this blog in June of 2005, it’s a concept I’ve been trying to hammer home.

And quite frankly, I can’t think of a better place to start this discussion than by quoting one of the first things I ever said on this blog about propellant transfer:

After quite a bit of discussion, I realized that your perspective on what constitutes “space development” or “space exploration” will greatly impact your opinion on how to best go about achieving those goals. In fact, this is a fairly valid general point: your focus determines your path.

If you think that having a small McMurdo-on-the-Moon style lunar science base is space development, then your opinion on what is an ideal method to accomplish that will differ quite a bit from someone like me who doesn’t see the moon as settled until there are dozens of settlements and hundreds of thousands of people living and working there.

If you’re only planning on sending a few people a year to a small government camp, with no intention of ever opening the moon up to commercial exploitation, building a big Shuttle Derived Heavy Lift Launch Vehicle may make a perverse amount of sense. If on the other hand you actually want to make money on the moon through say, lunar tourism, you realize that the only way you’ll ever get the costs low enough is if you have some sort of reusability built into your transportation system. Which means biting the bullet with on-orbit rendezvous, docking, propellant transfer, and vehicle assembly.

Ok, with that said, let’s jump into some of my thoughts on Rob’s article:

Missing the Elephant in the Room
In this article, Rob focuses on the economics of a propellant depot, and spends a lot of time trying to make the case that there’s no market for depots.  “But where are the customers?” he asks, all the while missing the most obvious customer for propellant depots: NASA’s manned exploration program.  I’m kind of confused really.  I’m not sure if this is some sort of a strawman–trying to make it look like we think NASA should go away entirely and that with propellant depots, commercial space can do all of the exploration by itself?  Because that’s not what any of us have been suggesting.  We’ve been suggesting that if NASA transitioned to an open exploration architecture, enabled by propellant depots, that it would allow them to do better exploration, while at the same time helping catalyze demand for commercial spaceflight and provide an anchor tenant for the infrastructure that we’ll need if we ever want to get past Apollo reruns.

I guess he could have merely forgotten the comments that started this whole conversation. As I understand it, this whole discussion started with a commenter to his post on space policy recommendations linking to Rand’s article on the same topic, which suggested ditching the NASA-operated HLV approach for a depot-centric commercial launch approach to space exploration.  Rand’s point if I may summarize was that if NASA opened up the lunar architecture to the concept of propellant transfer, not only could they launch the whole thing on existing EELVs (possibly in just two hardware launches), but they would also be opening up the largest launch market in history.  The demand for propellant on orbit for even a modest lunar program would be amazing compared to the current launch markets.

No Such Thing as a Free Launch
Before I go on about other depot markets, there were some other points in Rob’s article that bugged me a bit.  For example, one of them was when Rob goes on about all the costs of a depot.  But instead of listing anything new, he goes and lists a bunch of stuff that are typical costs for any spacecraft.  In fact, he lists several costs that are typically rolled into the cost of launch (like the cost of the launch’s ground support and the cost of shipping the rocket to its launch site).  And then he acts like this is some sort of major news that none of us depot supporters have ever thought of before.  Sure, propellant depots have costs associated with them.  We know that.  Nothing in life is free, but none of that is all that hard in the overall scheme of things.  What matters is whether or not you can keep your costs sufficiently low compared to your revenues to turn a good enough profit.  Unfortunately Rob doesn’t actually go into that at all.

LEO Satellite Demand for Depots
Getting back to markets for depots, I think Rob doesn’t really have a good idea of how a depot would likely be used.  It’s really easy to setup strawman markets that don’t make any sense, and then knock them down.  Sure, most satellites aren’t going to go “blast across to a new plane and dock for fuel.”  There are some LEO satellites that could actually use the added flexibility/lifetime that could be had if there were an easy way of refueling themselves on a regular basis (or achieving that goal through alternative means).  Rob dismisses these with the comment:

Telecoms and Earth observation satellite propellant resupply vehicles have been in development but to date, despite an obvious market and claims by developers of anchor purchasers waiting in the wings, nothing has materialised

I think Rob is missing several things here. First off, the capabilities necessary to do such services has only recently been demonstrated by the US. Most of the technology required for doing something like what Orbital Express did wasn’t really that far past what we had in the 1960s, in fact the Russians demonstrated autonomous docking and fuel transfer at around that time. The problem is that while this market is real–most of the people I’ve spoken with in industry agree that there is real need for the added operational responsiveness that tugs and depots could give LEO sats–especially after China’s recent ASAT tests–there are some serious obstacles standing in the way of potential suppliers of these services. For one, no existing US satellite is designed to be refueled on-orbit. There are ways of getting around this (you don’t have to refuel the target satellite per se for the satellite to benefit from the capability of prox ops tugs and orbital refueling), but that provides at least one hurdle that has to be met. Another hurdle is that while there is demand, a lot of it is for government satellites. And while the government can purchase off-the-shelf services, and can set up procurement programs, they can’t typically sign a contingency contract for future delivery. Which means that without the military actually procuring the money to run the development as a typical government contract, potential suppliers would have to build the systems on speculation. And the reality is that most of the big publicly traded aerospace companies have a hard time taking up projects of this size on speculation–it’s just too much risk, and most of them don’t have huge amounts of free cash to invest in such concepts either. On the opposite side of the scale, smaller entrepreneurial firms might be able to handle the risk (and maybe even raise the money), but they face the hurdle that the AF or NRO is unlikely going to let some no-name company get anywhere near one of their “precious national assets” without first making them go through a gauntlet of paperwork. It isn’t impossible, and there may very well be a way of getting around those obstacles, but they are real barriers to entry. But remember: barriers to entry do not mean that there is no market!

Lunar Robotic Explorers
Rob then goes on to unfairly dismiss another market, deep space robotic probes:

The question I raise is this, are there governments around the world that can pay to put together a 20,000kg robotic spacecraft that would also have docking technology (maybe they can buy that off the Russians) and would need the services of an in-orbit propellant depot?

Why do you need to have a 20,000kg robotic spacecraft to take advantage of a depot? Why does the spacecraft itself need to do the docking (or berthing)? The explanation Rob gives in comments is lacking. Right now, there is only one vehicle that can even deliver a payload of 15klb to lunar orbit (that’s about what Delta-IVH can do). If you wanted to send even a 16klb payload to lunar orbit, a depot would be very useful. But more to the point, Delta-IVH’s are extremely expensive. A single Atlas V 401 (or even Falcon IX) could launch the satellite and the transfer state (a refuleing capable upper stage), and then buy propellant from the depot. If the depot is commercial, and in an inclination where it can buy from the cheapest suppliers, you might be able to do such a mission at a far lower cost than you could launching it on a Delta-IVH. With a propellant depot, even stages as small as the Falcon I upper stage can put sizeable payloads into lunar orbit. A single Zenit launched tanker could provide enough propellant for something like five Falcon 1 missions. There’s no reason why customers would have to build super behemoth spacecraft to take advantage of propellant depots.

In fact, some countries, whose launchers are still fairly small (like India for instance) could benefit a lot from propellant depots. The GSLV could launch the biggest GEO sat on the market (minus circularization fuel), and in conjunction with a propellant depot, the upper stage could be refueled could then deliver those satellites to GEO. Without having to develop a new booster to do it. In fact a refueled GSLV could easily deliver even a spacecraft the size of LRO/LCROSS to lunar orbit.

Anyhow, propellant depots can make a lot of sense for launching planetary/lunar probes, especially if:

1. It allows you to launch your mission on smaller, cheaper launchers
2. The depot can buy propellant internationally from the lowest bidder (ie is commercial in a proper orbit)
3. It gives enough performance that your mission can be launched at all using your existing boosters (existing Atlas V Centaurs modified for refueling can deliver more than 3x what a Delta-IVH can put into lunar orbit without).

Anyhow, enough of that for now.  My key point here is that there are potential users for propellant depots.  Some of them (military LEO sats and possibly the Indian GSLV) that could benefit from the capability sooner rather than later.  Even if NASA decides to continue ignoring propellant depots as Rob suggests, it is a technology that can probably make its way to market.  But the path to commercializing those markets is going to be really tough without a solid anchor tenant.  If NASA did adopt a more depot-centric architecture, it could go a long way towards providing the demand and breathing room for propellant depot operators that would allow them to grow those other markets until they’re ready.  In other words, if NASA actually took seriously its legal mandate to “seek and encourage to the maximum extent possible the fullest commercial use of space”, it could make it a lot easier for propellant depots to come into existence sooner, and survive long enough for the markets to adjust to and adopt those new capabilities.

Heavy Lift
One last thought before I (belatedly) go to bed.  In comments to the post, Rob states that:

A sustained presence on the Moon would help but I still think that building bigger rockets is ultimately more efficient than the entire depot Earth to orbit industry, infrastructure you would have to build to deliver this propellant, that could be on-vehicle using one big booster.

That last bit is the whole point. With depots, you no longer have to fit it all on one booster, which now has to be massive and bloated to deliver even a tiny payload to the Moon. With depots, you have a lot more size flexibility. The same depot that one day allows you to send a small lunar probe into orbit without requiring a huge launcher may on a different day allow you to send a 20-person crew to the Moon or Mars.

There are challenges that will need to be overcome on the way to implementing commercial propellant depots, but they are truly worthwhile, because without them, we’re never going to become a spacefaring society.

For those of you who were there for the propellant depot panel that I chaired at Space Access this year, the paper covers in more detail many of the things that Frank Zegler presented.

After an introduction where the benefits of propellant depots for the planned Constellation architecture (such as allowing the architecture to actually, you know, work…), a concept for a first-generation propellant depot was given.  This concept was designed around some of the work they’ve done on their ACES stage (aka the Wide Body Centaur that I’ve written about previously), combined with some recent work on deployable sunshields.

ULA's Proposed Propellant Depot Concept

The paper hit on several of the key concepts that I’ve mentioned on this blog:

1. The benefit of “settled” cryogenic fluid management (CFM) instead of “zero-G” CFM.  To reiterate, if you can force the propellant to assume a preselected orientation, almost all CFM tasks go from being science projects to being straightforward adaptations of terrestrial CFM techniques.  Basically you want to keep liquid from going out the vent, and gas from being ingested into the transfer lines.  They propose a combination of the propulsive settling that they’ve demonstrated over almost 200 Centaur flights, combined with a rotational settling technique similar to what we’ve discussed on this blog in the past.  This rotational approach, and the transition to and from the axial settling to rotational settling is set to be demonstrated after the DMSP launch this next year.  They’ll have almost 11klb of unused propellants to play with after delivering the primary payload, and they plan to squeeze as many experiments as possible out of that excess propellant.  There’s another approach that Frank Z. and I were working on for an SBIR proposal a year ago that could potentially also work if this one doesn’t turn out.
2. Proper thermal design can allow for passive systems that minimize or eliminate boiloff.  While you can add an active cooling system to compensate for a poor passive thermal design, it’s much better to do what you can first with a good passive design.
3. Almost all of the technologies for propellant depots are already developed, many of them to high TRLs.  Especially if you go with settled cryo handling instead of insisting on zero-G.
4. If NASA opened up its lunar architecture to allow for the use of propellant depots, it would greatly expand the current demand for orbital launches.  As the authors point out, even just topping-off the Earth Departure Stage’s LOX tanks would provide something like 10x the mass demand as COTS will.
5. They also discussed the importance of having experimental facilities for flight testing and maturation of these technologies before they’re implemented on real systems.  They mention their Centaur Test Bed concept for cryogenic experiments as secondary payloads on Atlas V, but they also link to an interesting paper by Dr. Chato of Glenn Research Center about the history of suborbital and orbital flight testing of CFM technologies.  I think this is one of those areas of research for which a low-cost, unmanned suborbital vehicle like we’re developing at Masten could greatly aid the development and maturation of critical spacefairing technologies.

There were a few issues I had with their presented concept that are probably worth mentioning.  First, they focus on only providing LOX.  While this may still be useful for NASA missions, it’s not as useful for commercial missions.  Since hydrogen boils off a lot faster than LOX, not having a way to top off your LH2 tank on orbit eliminates one of the big benefits of propellant depots.  Even if you don’t go with LH2 for your fuel, having both oxidizer and fuel at the depot gives you far more flexibility than just the one fluid.  Of course, this disagreement is mostly just a matter of taste on my part.

My other concern is more substantial.  They only briefly mention this in the paper, but the sunshields they’ve been working with use aluminized plastics.  Unfortunately, the LEO environment is somewhat nasty on plastics due to atomic oxygen.  In order to minimize degradation of their sunshield, as well as minimizing damage from space debris, they selected a 1300km altitude for their analysis.  While this makes the sunshield work better, that altitude is not a great place for a depot operationally.  First off, it’s inside the edges of the inner Van Allen belt.  Once you get much higher than about 500-600km, the radiation dosage goes way up.  This makes it a lot trickier on the electronics, and I don’t know if you could keep the rendezvous, docking, and transfer periods short enough in such an environment to avoid radiation damage to the crew (the point of this depot after all was for providing propellant for crewed lunar missions).  Ideally, a propellant depot should be a place where you can loiter for a while in case something comes up that delays the mission.  Lastly, 1300km is high enough up that there’s a significant penalty for delivery to that altitude.  Especially for future potential RLVs.  Now of course, a tug could relax that constraint a bit (and as I mentioned in my previous post would make operations a lot better in general).  But I think the reality is that in order to close this case operationally, they really need to find a way to make the sunshield survivable at lower altitudes.  For non-LEO propellant depots (L1/L2, LLO, Mars Orbit, etc) this shouldn’t be a problem, and the idea can probably be used as-is.  But the concept probably needs some rethinking if they can’t get it to work at a reasonable orbital altitude.

[Update:  I was able to dig up a bit of additional information about the concept.  Apparently the 1300km number was somewhat arbitrary.  The concept can work at more reasonable altitudes (ie 400km would probably be fine), it's just a question of how long of a lifetime you want for the sunshield.  With the petals concept as Frank explains it in the comments section, it sounds like down the road you might be able to replace the sunshield if it wears out, but depending on the lifetime, it may make more sense to just retire the module at that point and launch a new one.  Basically, it sounds like a tradeoff between lower altitudes for easier access vs. more maintenance/replacement costs due to more wear.  But this information more or less ends that key concern of mine with the concept.]

One last thought is that ULA is still pulling its punches on this technology.  They talk about how it could help aid the existing Constellation architecture, but the reality is that once you have this technology, you could completely transform the Constellation architecture, or get rid of large chunks of it entirely.  Once you have propellant depots you no longer nead super heavy lifters like Ares V.  Depots allow you to store the propellants you need for long durations, so that the ESAS concerns about losing a mission if a given launch is delayed or failed are greatly reduced.  Depots allow you to split propellant launches up over as many redundant launchers as you want.  If you look carefully, you’ll notice that the ACES stages they mention at the end of the paper could carry quite a bit more propellant if you have a Depot to top them off than an un-topped EDS stage.  And if you launch that stage dry, you can have a system that has better cryo thermal properties, much better performance overall, and it would be part of a system that was commercially useful for other markets.  Once you go to a propellant depot architecture, you could launch all of the actual dry hardware from the ESAS architecture on two existing or near-term EELV Heavies, and then the rest of your launches you really don’t care about launcher reliability.  Basically with a propellant depot architecture, you can keep the number of mission-critical rendezvous and docking opportunities to the same number as ESAS, while greatly increasing performance, reducing cost, and stimulating the private launch industry.

Like Space Tugs, propellant depots are an idea whose time has come.

One of the main ideas presented in the paper is a “Payload Bay Fairing” that would allow a heavy EELV to interface and launch payloads originally designed to launch on shuttle. The EELV would deliver the PBF with its encapsulated payload to just outside the ISS “visiting vehicle stay-out zone”, and then a tug of some sort would provide “last mile” services, hauling the PBF and its payload to the station, where it would be unloaded. They mentioned using Soyuz/Progress as the tug (like Constellation Services proposed), but decided to focus on ATV due to concerns about ITAR and INKSNA issues.

Payload Bay Fairing© (courtesy United Launch Alliance)

The PBF would be derived from the current 5m payload fairing used on the Atlas V. The PBF would have docking/berthing mechanisms on both ends, and would have structure to allow it to transfer loads into the payloads in a manner similar to the shuttle payload bay. I imagine it would also provide the necessary services for maintaining those payloads until they were ready to be installed at the ISS. By launching this on Delta-IV, you could pretty much deliver any payloads that the Shuttle was supposed to deliver, and probably at a far cheaper price. This includes the MPLMs, the AMS module that has received so much attention, and any other modules that isn’t going to get a shot at getting launched due to the 2010 shuttle retirement.

This is an interesting idea in several ways:

1. For small payloads that the COTS providers can deliver, a Delta-IV/ATV derived solution isn’t going to be cost competitive so it doesn’t necessarily step on toes as much.  If the Shuttle is kept running after 2010, being a government jobs program it doesn’t have to be economically competitive with COTS, and therefore could easily squash the nascent efforts by SpaceX, OSC, and others in this area.  With a Delta-IV based system, procurements would have to be handled in a competitive manner if the payload is one that could be flown on other commercial options, and therefore it’s much less likely to interfere in that key initial market.
2. This provides a commercial method for replacing the key functionality that we’ll be losing when the space shuttle retires.  This might allow us to drop the albatross sooner.  More importantly it might allow for some of the other modules that were deselected to be restarted and launched.  If Atlas V ever gets built there would even be some redundancy.  Building a station out of 20 tonne chunks isn’t a crazy idea so long as all those chunks aren’t stuck flying on the same system.
3. Space Tugs for proximity ops are an idea whose time has come.  If you start with an ATV-based tug system, that might provide enough of a market for other more affordable competitors to start filling that niche.  Once you have space prox-ops tugs available, lots of things become much, much easier.  Most of the mass launched with Shuttle or Progress (or even ATV or HTV) ends up being used to handle things like prox-ops, rendezvous and docking, cargo handling, reentry, etc.  The more of those functions can be offloaded to something that can stay in orbit and not have to be relaunched every time, the higher a percentage of your delivered mass can actually go to paying cargo, propellants, or passengers.  Also, by removing offloading a lot of the Visiting Vehicles requirements to the tug, it makes it removes a big barrier to entry by new suppliers.Space tugs would also benefit people like Bigelow. If he didn’t have to design each of his modules as maneuvering, independently operable spacecraft, I bet his task would be a lot easier. Also, tugs would make it much easier for different groups that want to dock/berth with his stations to do so.

It’s also amusing to note that I tossed out such an idea on usenet back in 2003 right after Columbia.  Now, I’ll admit that at the time, I really had no clue of all the challenges involved.  And accusations that I sounded like an “engineering undergrad with lots of imagination but very little experience with real world considerations?” were probably more accurate than I’d like to admit now. But it’s always cool finding out that one of my ideas I had years ago actually was a good one after all.

• The report mentioned that our ISS experience shows the importance of having redundant transportation methods (ie imagine what would’ve happened to ISS if Soyuz didn’t exist). I don’t think that redundant transportation method should necessarily be another government-centric transportation system, but I agree wholeheartedly that monocultures are a bad idea.
• The report also mentioned that having a safe-haven in LLO is one of the best ways to increase the safety and flexibility of a lunar exploration program. Right now, most of the danger associated with lunar exploration have to do with operations on or near the moon. The current architecture does nothing to reduce those risks, but instead focuses on the much sexier earth-to-orbit transportation risks. Having some infrastructure in LLO can go a long way to fixing that, while also giving you some very interesting mission options. Now, I’m still a fan of the idea of Lagrange stations, and I think that in the long-run they’ll dominate the traffic in the lunar half of cislunar space. I just think that there is a small, and critical niche filled by one or more small polar LLO stations. I’ve been planning to write up my ideas on this concept for over two months now, so can someone poke me in a few weeks if I haven’t followed up on this thought?
• Unlike NASA they don’t seem to be deathly afraid of on-orbit assembly when it makes sense. Of course, they don’t have an HLV fetish that they have to rationalize…

There were a few other good points, but those three were the key ones that stood out to me. Of course they also seem to be missing the importance of propellant transfer, and they seem to be almost as clueless as NASA as far as commercial enterprise is concerned (both why it’s important, and how best to foster real commercial involvment). But it was an interesting read if you have a few minutes.