DARPA wants to cut the risk - and the cost - from servicing satellites in space

Graham Warwick/WASHINGTON DC

4877

Once in orbit, satellites are essentially unreachable. There have been rare, and extremely expensive, examples of spacecraft being retrieved and repaired in orbit or returned to earth, but basically, once it is launched, a satellite is on its own.

4878

Now the US Defense Advanced Research Projects Agency (DARPA) wants to change a paradigm that has been fundamental to spacecraft design since Sputnik: that a satellite must be launched with all the fuel on board it will ever need. "It's like buying a car and driving it out of the showroom towing a tanker with all the fuel it will ever use," says Sam Wilson, programme manager for DARPA's Orbital Express technology demonstration.

Orbital Express will be an on-orbit demonstration of technology for not only in-space refuelling but also replacement of electronic components - "plug and stay", as Wilson calls it. The aim is to make spacecraft more like aircraft, which can be refuelled in the air and updated over their service lives.

The advantages of in-space refuelling are potentially enormous. If satellites could be refuelled in orbit, they could be launched with smaller fuel tanks, saving cost. Fuel which now has to be used sparingly could be expended to manoeuvre the satellite in orbit.

For the military, the ability to manoeuvre in orbit has several benefits. Wilson says: "If a satellite could accelerate, decelerate, change orbit or altitude, it would not be so easy to track. It could show up early and stay longer, making it difficult for the enemy to plan." Reconnaissance satellite orbits could be adjusted to cover new trouble spots. "We could make do with fewer satellites - instead of 24 maybe 16 or less - so we would save money by manoeuvring."

Equally appealing is the ability to update spacecraft in orbit. "It takes a long time to launch a satellite," says Wilson. "They're obsolete when launched and then they stay up there for five to 10 years. It would be great to be able to add modules and update the brains to get a more capable satellite."

The combination of in-space refuelling and on-orbit upgrading would extend satellite service life and allow fewer, smaller spacecraft to perform the same mission. A 24-year mission that currently requires three 6,000kg (13,200lb) satellites with eight-year service lives could be performed by two 3,200kg satellites with in-space refuelling and 12-year lives, Wilson says.

To demonstrate the required technology, Orbital Express has two main elements. The first is the ASTRO - Autonomous Space Transfer and Robotic Orbiter. "This lives in space," says Wilson. "We send up fuel and electronic modules separately. ASTRO captures them and takes them to the satellite - with no people involved."

Today's spacecraft are not designed to be refuelled and upgraded in orbit, so the second element of the programme is the NEXTSat - a satellite representative of the next generation of spacecraft designed to be serviced by the ASTRO. During the Orbital Express flight demonstration, planned for 2004, the ASTRO will be required to dock with the NEXTSat and transfer fluids and modules "multiple times", Wilson says.

Water rocket

In addition to the Orbital Express demonstration, DARPA is looking at other technologies which could enable on-orbit servicing. First among these ancillary technologies is what Wilson calls the "water rocket".

"What is the right fluid to transfer?" he asks. According to industry responses to a DARPA request for information the answer appears to be water. As a result, the agency hopes to fund a parallel demonstration programme.

Using water as the fuel on future satellites has several potential advantages. First, the spacecraft battery can be replaced by a reversible fuel cell. "Solar power runs the satellite by day and excess electricity breaks the water into hydrogen and oxygen, which is stored. At night, we mix the hydrogen and oxygen in a fuel cell, generate electricity and regenerate the water. It's a reversible process."

Fuel cells have particular attractions for low earth orbit (LEO) satellites. "LEOs are in shadow 16 times a day. Cycle a battery 16 times and it wears out - 40,000 to 50,000 cycles over the life of a battery is a lot."

Water can also propel the satellite, in several ways, Wilson says. Gaseous hydrogen and oxygen can be burned in a chemical rocket. Alternatively, electricity and water can be combined in a Hall effect or steam thruster with higher specific impulse but lower thrust than a chemical rocket. "It won't move fast, but will be more efficient," says Wilson. Finally, gaseous oxygen could be used in a cold gas thruster suitable for proximity operations where the exhaust could impinge on the satellite being serviced.

There are other potential advantages to using water as the fluid of choice for orbital operations. "If all we are launching is water, we can use higher risk, lower cost launch systems such as guns," says Wilson. "Costs could be an order of magnitude less, allowing us to launch fuel cheaply. We could launch water in bulk and set up fuel farms in space".

Even at today's launch costs, Wilson says, enabling satellites to stay in orbit longer will save money. Typically a 2,000kg satellite will carry 500kg of fuel, enough for eight years of operation. A refuellable satellite launched with fuel for only the first two years of operation would weigh around1,600kg, he says, reducing the launch cost. Additionally, the satellite would stay in orbit longer, "saving the price of launching another satellite."

Keeping satellites in orbit will raise other issues, as current spacecraft are designed for relatively short lives. Part of the Orbital Express programme will be studies to look at the effects of longer in-space service lives on components such as solar arrays.

DARPA has awarded concept definition contracts to three teams, led by Boeing, Lockheed Martin Sanders and Spectrum Astro. The make-up of the team is important, Wilson says, as one of the key products of the Orbital Express programme is intended to be industry-standard interfaces for the docking and transfer of fluids and modules between spacecraft.

Each of the three teams includes at least two satellite manufacturers. Boeing's team includes the former Hughes Space and Communications (now Boeing Satellite Systems) as well as Ball Aerospace and Technologies and TRW. Sanders is teamed with Lockheed Martin Space Systems and Space Systems/Loral, as well as Charles Stark Draper Laboratories and Moog, while SpectrumAstro's team includes Ball Aerospace, as well as Science Applications International and Oceaneering Space Systems (Flight International, 10-16 October).

On the winning team "one guy will build NEXTSat and one will build ASTRO", says Wilson. "That way they are forced to make sure they can interface, and don't come up with a contractor-unique solution."

GEO too far?

Over the first three months of the 14-month first phase, the teams will look at potential missions that could benefit from in-space servicing. "DARPA's first cut is that LEOs would benefit most because of the battery cycling issue, which is not a problem with GEOs [geostationary orbit satellites]," Wilson says. "Also, getting fuel to a GEO could be a problem."

The agency plans to combine the studies from all three teams, and produce a composite mission around which the contractors will then design their proposed demonstrations. After completion of the first phase, DARPA plans to select one team to build and launch the ASTRO and NEXTSat and conduct the demonstration.

Also, after the first four months of the programme, DARPA plans to focus the contractors on a single set of interface standards to be used in designing the demonstration spacecraft. "If Team A has the best fluid interface and Team B the best electronic interface, then Teams A, B and C will use them," Wilson says.

DARPA expects the on-orbit demonstration to be representative of more than one potential mission scenario. Both the ASTRO and NEXTSat must be representative of production vehicles, but contractors can modify an existing satellite bus for the NEXTSat provided the ASTRO is scaled accordingly.

"The scaled sizes are important," says Wilson. Satellites behave differently docked than they do when flying separately, because the centre of gravity of the docked complex is different. "Contractors will have to decide how to control the bigger mass - using ASTRO or NEXTSat or both. Will the satellite have to stop working while being refuelled or will it be able to continue its job?"

Other issues to be resolved include how the ASTRO will navigate itself in orbit. If the mission involves docking with a LEO spacecraft, the orbiter will be able to use global position system (GPS) satellites in medium earth orbit, but if the mission is in GEO, GPS will not be available, Wilson says. Another involves the type of robotic vision sensor to be used by the orbital for docking, as this will have to work in both daylight and darkness.

Wilson details one possible demonstration scenario in which the ASTRO would dock with one face of the NEXTSat, transfer fluid, then undock, manoeuvre round to another face, then dock and transfer electronic modules. The faces of the NEXTSat could be representative of different types of satellite and different interfaces, he says. One face could have an antenna farm that the ASTRO would have to manoeuvre around while another could have clusters of thrusters which the robotic orbiter would have to stay clear of. The NEXTSat could also be manoeuvred in space to represent different types of satellite stabilisation. "With gravity-stabilised satellites the ASTRO would have to dock on top," Wilson says.

Robotic docking

There have been previous demonstrations of robotic docking in orbit. In 1998 and 1999 Japan's National Space Development Agency successfully accomplished the automatic docking of two satellites - one small and one large - using its ETS-VII engineering test satellite. The test flight also involved the use of a robotic arm, tele-operated from the ground.

DARPA plans to go much further. The goal is for the ASTRO to be autonomous, says Wilson. "It's the only way to be cost-effective." Also the docking could take place out of range of ground stations, ruling out tele-operation, or in GEO, "where the speed of light is an issue", he says. The ASTRO will deal with co-operative satellites and "will only dock with things it knows", simplifying the task. Wilson says: "If it's fully automatic, we can make it affordable."

Affordability is a major factor in the Orbital Express programme. "In-space servicing has been studied for 30 years, and every study says it pays off. But when they built in the risk, they always decided not to do it," says Wilson. "Orbital Express will take the risk out and demonstrate we can do it affordably."

The end product of the demonstration will not be a vehicle that can immediately be put to work in orbit. "At the end of the programme we will have a set of non-proprietary interface standards that can be adopted by industry," says Wilson. "This is the programme that will create those standards."

Development of the ASTRO concept beyond a technology demonstration will be the responsibility of the first satellite programme that decides it wants an on-orbit servicing capability. "They will do the engineering and manufacturing development on the ASTRO, the fuel farms, etc," Wilson says.

DARPA anticipates robotic orbiters will be developed to service specific satellite constellations, or spacecraft within a specified range of orbital inclinations and altitudes, rather than being roving repairmen. But the vehicle will live with the satellites it services, rather than being launched on demand. "One of ASTRO's powers is that it lives up there," Wilson says.

4879

Source: Flight International