Orbiting Supply Depots

by Tom Hill

Let's start with the standard cliché: Getting into space is expensive. In the 46 years since the first time someone took a ballistic missile, strapped a satellite to it, and sent it to orbit our globe, this statement has been accepted as truth. Several methods have been proposed to reduce the cost of flying into space; some have been built. Depending on the accounting methods used, some of the flying ones have achieved their goal, but the cliché stands.

Arguments abound as to why going into space remains expensive, but I feel that they boil down to a couple of facts: 1. We don't do it very often, and 2. There's very little usable material in place when we get there, so we have to take everything with us. An orbital supply depot may solve both problems.

Current Impasse

To illustrate the first point, let's take a look at how many flights into space happen each year. In 2001, there were 58 launches. Launches per year have been dropping since the 1975-79 timeframe, when on average 121 launches took place each year (Statistical Abstract of the US, 2002 chart #778). These numbers are very low compared to other levels of human activity. Each day in the year 2000, an average of 24,635 (Statistical Abstract of the US, 2002 chart #1035) aircraft departures took place in American airspace, and during that same time period, millions of cars ran, carrying people with varying efficiency where they needed to go with varying urgency.

Even though the overall space activity seems low, when divided by the number of parties doing the activity, the numbers get even worse. Today, there are a dozen launch service providers vying for this year's flights. That translates to very few missions for each provider. In fact, the numbers are so low that the launch industry barely qualifies as an industry. When a few craftspeople hold the arcane knowledge required to do amazing things, we're talking more along the lines of a guild or some other medieval term that barely finds use today.

Because of the expensive practice of space launch, items that fly in space are designed to maintain longer rather than shorter operational lifetimes. Satellites that provide our communications regularly last ten years or more, and other satellites follow suit. The only activity with relatively short timelines is human spaceflight, where supply runs to keep people fed and hydrated are high priority for health reasons, and construction flights are high priority for political reasons. In the case of human spaceflight, concerns such as 'man-rating' a craft also work to keep flight rates low.

Yet low flight rates contribute to high expense per flight. Imagine how much it would cost to fly commercially if only 5 of each type of aircraft (747, A320, etc) were built each year, and only flown once? The same can be said for cars, though with the cost of some modern SUV's, one has to wonder whether they're produced with a space-launch mindset. Air travel and car travel became cheap because there was a demand for the activity, and that demand drove new manufacturing and operating techniques that allowed costs to drop. Right now, in space, we've grown accustomed to low flight rates, so vehicles that take people and things there are built one-at-a-time, practically by hand, and the objects that fly there are designed to last long enough to perpetuate the low flight rate.

So, how do we break out of this?

The Plan

Here, the goal is to provide a reason for frequent flights into Low Earth Orbit (LEO). To make sure that the flights are not frivolous, each is tasked with taking a significant amount of raw material (water, in this case) to a destination in orbit. By allowing an arbitrarily high flight rate, with relatively low costs for the loss of a payload, a startup company can build its expertise in rapid-launch rocketry. After flying a number of missions (let's say one a week for a year), they have a sound statistical case for claiming that their vehicle is safe to fly. Then, that vehicle could be used for other applications in LEO such as space tourism. But first, we need a destination.

An Orbiting Supply Depot

Imagine a place in Earth orbit where any craft in that orbit could stop and pick up hydrogen for use in rockets or fuel cells, water for its crew, and oxygen for multiple uses. For missions to LEO, these supplies could cut the need for frequent logistics flights from Earth's surface. For missions beyond LEO, these supplies could make such journeys feasible using the relatively modest-sized launchers that we have today.

The entire system can be described as an orbiting electrolysis station, where moderate (10,000 kg class) payloads of water are delivered and converted into liquid hydrogen and oxygen. Since water can be stored on orbit with relatively little maintenance, there is no penalty for a high flight rate, as the water can simply be held in place until the electrolysis process catches up. Visitors can then take on the supplies, as required. This type of station has been proposed before (Smitherman, et al, CO School of Mines, 2001), but its implementation was seen as being feasible after some sort of mass driver exists. I assert that waiting for a mass driver moves the propellant depot into a realm of super-science which is unnecessary to make a supply depot useful.

The Depot

The depot is designed to fly in a single flight of a heavy-class payload by today's standards, which translates to Evolved Expendable Launch Vehicle, Shuttle, or Ariane V. If constrained to a shuttle launch by either politics or deployment complexity, the cost numbers discussed later will have to be increased. The depot is pictured - deployed - in model form in Figure 1. At one end of the depot reside the docking units for water deliveries. The system is designed to allow multiple dockings at once. From these docking locations, the water flows into an electrolysis unit, which splits the water into its components: hydrogen and oxygen. From there, the component gases move to a refrigeration system that cools and pressurizes them into their liquid form. The cryogens are then forced into tanks for storage by the follow-on products.

At the far end of the depot rests the docking unit for customers, as well as a set of thrusters to provide acceleration as needed to make all the fluids flow in the proper direction. Acceleration along a primary axis allows a simple flow method that doesn't require pumps, bladders, or any of the other contraptions that come with moving liquids in zero G, providing for a much simpler system. The docking unit allows access to the liquid hydrogen and liquid oxygen from the cryogenic tanks, as well as to the water from the original deliveries.

My original goal in this design was to stabilize the depot passively, though as the idea matures, that becomes less likely. Because of the huge power requirements for both the electrolysis and the cryogenation process, any solar arrays supporting the mission would have to be larger than that of the ISS. By my estimation, it would be best to hold them steady and pointed at the sun by turning the entire depot. This has the side benefit of keeping the cryogenic tanks in the solar arrays' shade.

If the system catches on, and customers need supplies at a greater rate than one depot can produce or store, expansion is a possibility. By docking an additional depot at the customer end, it's possible for the first depot to double its storage capacity. Increasing production rate is another matter, because more power will be required, and that would be difficult to accomplish with solar power, considering that the entire second depot would be shaded by the first depot's arrays.

I foresee the depot being built as a typical, government procurement, with all the trappings of budget overruns, congressional hearings and news-show installments. Once completed and on orbit, it will be time to pay someone to deliver water to the depot using capsules built largely the same way.

The Delivery Capsule

The capsule (pictured in figure 2) that carries water to the depot will be designed around a tank with docking ports at each end. Being dual-ended allows capsules to dock either to the depot, if they're the first to arrive, or another capsule if it arrives later. The tank will be insulated and heated as necessary to keep the water onboard from freezing. Power for the heaters will come from the depot when the tank is docked. The tanks are designed to handle pressure, as waste gas from the fuel depot will be used to pressurize them, and allow a rapid flow of water into the system. As tanks empty, the same pressure will force them to disconnect from the depot when the tank is unlocked.

Delivery to the depot from the injection orbit, as well as proximity and docking maneuvers, are handled by the orbit assist ring or OAR. The OAR provides enough propulsive power to move the capsule to a docking with the depot, keeping the water warm as necessary. Once the capsule is in position, the OAR separates, and de-orbits to burn up in the atmosphere.

Depending on the method chosen to liquefy the hydrogen and oxygen, it may be necessary to have additional cargo vehicles bring liquid helium to the depot. This complicates the system somewhat, but the trade may pay off in a simpler, longer-lasting supply depot. It would also increase the flight rate necessary, if helium were expended in the process.

In this discussion, the lift mass of the water is more important than the lift mass of the delivery capsule, but the capsule's mass is non-zero, so it needs to be discussed. Estimates on the mass of the capsule and the OAR range from 15-20% of the cargo carried, so any discussion of launching 10,000 kg of water into orbit will need to include 2,000 kg for the tank and OAR.

The Launch Vehicle

In this concept, the design of the launch vehicle is of no consequence to the outcome. Because of the low payoff per kilogram offered and the high flight rate, it's likely that the vehicle and its support operations will have to be radically different than any method so far devised to fly into LEO. I believe that a significant portion of the vehicle will have to be reusable to prevent the complete manufacturing of each one for flight. Because of the high cost of labor in the aerospace industry, it's likely that the production and launch teams for this vehicle will be small. By taking both actions, the primary flight cost will be propellants, not wages. What makes this transition possible is payment upon delivery of the water, not payment for booster development, as is the norm today.

The delivery-on-orbit contract is becoming more common in government space activities, with examples including the Navy Ultra-high Frequency Follow-On (UFO) communications satellite and the Geostationary Operational Environmental Satellite (GOES) series due to fly in the next years. The version of this method envisioned here would set a new precedent. Previous delivery-on-orbit missions were small scale, on the order of 5 total launches. These kinds of numbers never allowed a real cost savings because by the time a team got used to the delivery-on-orbit approach, the mission was over.

For this method of delivery-on-orbit, the numbers are much larger. After choosing a payload size (10,000 kg used here), a dollar amount for that launch is chosen. For this example, we'll use $10M per launch, yielding an unheard-of per-kilogram rate of $1,000. A flight rate must also be chosen. Here, I'm partial to one a week for a year, because it's an unprecedented flight rate in modern times, and any vehicle that can fly once a week should be able to fly more often, as well. If this rate is chosen and guaranteed for one year of flights, it would provide an incentive of $520M to a company who has a 100% success rate on their vehicles.

The Payoff

At the end of the initial round of 52 flights, figuring a 90% overall success rate, there would be 468,000 kg of supplies in LEO, stored as either water, hydrogen or oxygen. This is an unprecedented amount of mass, and it would allow a lot of actions that were unthinkable before. Assuming a total cost of $1B (see cost estimates in the next section), the dollars spent to get a kilogram of this useful material into orbit is around $2,150. If no one else has adopted the cargo launcher as theirs, and flights are still taking place using current expendable rockets and/or expensive shuttles/replacements, these supplies can be sold at a profit in LEO.

If, on the other hand, the booster used to loft water to the depot becomes the booster of choice (a pretty big assumption, involving true market forces coming into play in the space industry, but work with it), then anyone using the low-cost booster need not purchase supplies on orbit. Here, the goal of low-cost flight to LEO has been achieved, and the supplies at our depot can be used for deep space exploration. For example, the Atlas III booster can launch 8,610 kg into LEO using its Centaur upper stage. Assuming that the Centaur were upgraded to allow precision maneuvers and if that stage were refueled in orbit then allowed to fire again, it could send its payload on the same trajectory that the Cassini mission traveled in its journey to Saturn. The costs of flying on a Titan IV (what Cassini actually flew on) vs. an Atlas would be the maximum savings recorded using the depot. In another example that I find more exciting, the Mars Direct mission architecture for sending people to Mars requires two launches of 140 tonnes into LEO. About 100 tonnes of that LEO mass is propellant. So, with a propellant depot on orbit, only 40 tonnes of crew habitat and empty tankage would have to be launched from Earth, which is feasible with the boosters of today.

The Cost

I believe that this plan can be implemented for a cost on the order of $1B. This includes 52 delivery missions to orbit at $10M each, building the supply depot for $250M, launching the depot on a heavy EELV for $150M (a planning number - individual flights are negotiated), and building 52 delivery tanks for $100M. This is not a small amount of money, but given the payoffs already discussed, it is money much better spent than that thrown down a dead-end research project.

The costs mentioned above are only for a successful project. If, for some reason, no carriers rise to the challenge of taking deliveries of water to an orbiting depot, the project can be abandoned at a significantly lower cost ($350M assuming that the depot and all of the delivery tanks are built) without launching anything. In the case of a partially successful project, only deliveries are paid for, so there are fewer supplies on orbit but at a significantly reduced cost.

It's possible that $10M for 10,000 kg is a completely unrealistic cost for launch to orbit. If this turns out to be the case, and the goal is still seen as valid, the price can be raised to provide further enticement.

On the other hand, there may be more interest in meeting the challenge. If more than one company builds a viable vehicle (demonstrated through flights, of course, not just presentations) additional carriers and depots may be built, all the while increasing our resources in orbit.

A Demonstrator Mission

A lot of things required for this mission are relatively common on Earth, but haven't been done in space. A demonstrator mission that flies an electrolyzer and cryogenic cooler would go a long way to answering some of these questions. A rather simple demonstrator payload could fly on board the Shuttle or Station if it meets safety requirements, but this demonstrator would have to rely on some sort of microgravity system pushing fluids through. If the mission flies on its own, it would provide a more rigorous test of the system, including the simplifying acceleration/pressure feed described earlier. The freeflyer demonstrator mission would be much more complex and expensive, but would model the idea with greater fidelity.

The Issues

Unfortunately, this system is not ready to go, out-of-the-box. There are issues - from the technical to those dealing with established attitudes - that must be overcome before this concept can become reality.

The first is power. This depot would require a huge amount of power, first to electrolyze the water, and second to cool and compress the products. Even with unprecedented power supplies (100 kW) devoted solely to electrolysis, the water processing rate would be only 8.18 kg per hour using an existing ground-based system. At this rate, it would take 51 days to process a 10,000 kg load of water. Of course, the electrolysis process is very scalable, so more power would allow a greater flow rate. The type of power to use (solar vs. nuclear) is a favorite topic for some, and if nuclear power becomes more acceptable through NASA's Nuclear Initiative, it may be an option, but that's a discussion for another day.

Once the products are generated, storing them is not easy. Hydrogen is especially difficult to keep, due to its low density and notorious tendency to leak. Adding additional depots below the first can solve the volume problem, but this won't be practical until the concept has proven itself useful. Leaks in valves and pipe joints will have to be minimized through meticulous construction.

The supply depot will be one of the oddest spacecraft ever built, as far as attitude control. It will start out relatively light, with large solar arrays and the majority of its weight on one side of them. As water deliveries accumulate, the center of mass on board will shift radically, only to be offset slightly by processing through the electrolyzer.

In real estate, the three most important words are location, location, location. In orbit, this translates to inclination, inclination, inclination. Where should an orbiting supply depot go in orbit? The answer is not intuitively obvious. The space station is inclined at 52 degrees to the Earth's equator, but the most efficient launches from The United States take place at around 28 degrees. Any corporate interest worth its salt would want the largest advantage possible in its launch, and base operations near the equator. This will require some serious marketing analysis, the likes of which space activities have never seen before.

The fact that nothing like this has ever been done is always a momentum barrier. A good answer is, "That's OK, low-cost launch to orbit hasn't been done either; maybe they match." Another philosophical point that needs to be addressed is that this system is biased towards hydrogen-oxygen systems. It's true that the electrolysis process produces those propellants in excellent proportion for an H2/O2 engine, with a convenient remainder of oxygen for use as breathing air, but a nuclear-thermal rocket could easily stop by for a fill-up, taking only hydrogen for its engine. If the payload for this rocket were a crewed capsule, they'd still be able to use the oxygen and water that the system supplies. Leftover oxygen could be sold to orbital stations with other crews on board. Let's face it; humans are addicted to oxygen, and without some major genetic alteration, that isn't going to change just because we have stations in orbit.

A Final Thought

Humankind has painted itself into a corner in space flight. The cost of flying to space is too high to do it often on today's budgets, yet we need to do it often to lower costs. If we're going to break out of it, something radical has to happen, either in the form of an external event (asteroid spotted on its way to us), competition (race to Mars?) or a small-scale mindset change that can grow. I believe that orbiting supply depots provide the latter, all the while building our on-orbit infrastructure for the future.

Tom Hill works in the aerospace industry on weather satellites, but his passion is getting humans into space to stay. He's published a book detailing the past, present and possible futures for humankind beyond the atmosphere titled Space: What Now? (ISBN 1-4137-2808-1 available at either Amazon.com or PublishAmerica.com), and can be contacted at hillkid@earthlink.net.