The Aug. 31 launch of the Palapa D telecommunications satellite by a Chinese Long March 3B rocket left the satellite in a badly off-target orbit and presented satellite builder ThalesAlenia Space with a problem: What should we do now?

Guiding the satellite into a destructive re-entry over the Pacific Ocean was quickly discarded as a scenario. A defect in one of the rocket’s upper-stage engines had placed the satellite into an orbit that, while misshapen, was not so bad as to envisage a total loss.

A decision on how to bring the satellite into a recoverable position needed to be taken immediately. Cannes, France-based ThalesAlenia Space in this case was not only the satellite’s manufacturer, but also the customer for the Chinese launch and the owner of the $200 million insurance policy as part of what is called an in-orbit delivery contract.

But Palapa D’s ultimate owner, PT Indosat of Indonesia, had to be consulted as to how to proceed, as did the insurance underwriters. Decisions needed to be made in a matter of hours.

In the following days, ThalesAlenia Space performed a series of maneuvers to stabilize the orbit. Palapa D is expected to arrive at its operating position at 113 degrees east by late October. ThalesAlenia Space has filed a partial insurance claim, telling underwriters that the satellite will have full use of its payload for at least a decade. That’s good news for PT Indosat since it will have to replace Palapa C2, the satellite now at that orbital slot, in 2011.

Michel Fiat, ThalesAlenia Space’s chief technical officer, spoke with Space News staff writer Peter B. de Selding about what it took to salvage what first appeared to be a desperate situation.


What is your current estimate of the useful life of Palapa D?


We are very comfortable in saying it will have at least 10 years of service, meaning enough fuel to assure more than 10 years in stabilized orbit, before considering any inclined-orbit operations.


Where was the satellite supposed to be dropped off in orbit, and where was it actually placed?

The satellite was supposed to be placed into what is called a super-geostationary transfer orbit with an apogee of 50,291 kilometers and a perigee of 200 kilometers, with an inclination relative to the equator of 21 degrees.

The higher apogee makes it easier to remove the inclination, to near zero, in orbital maneuvers after launch, than if we were placed in an orbit whose apogee was 36,000 kilometers  the classic geostationary transfer position. The Long March 3B is a powerful rocket, and it has no trouble placing a satellite the size of Palapa D into the super-geo transfer orbit. It helps us minimize the use of fuel as we move the satellite into final operating position in geostationary orbit.

In fact, because of the launcher under-performance, Palapa D was placed into an orbit with an apogee of 21,223 kilometers, and a perigee of 218 kilometers — so the apogee was substantially lower than what it should have been. The inclination was 22.4 degrees, more than 1 degree higher than the target.


Did you realize immediately that the drop-off point was way off nominal?

We did not have tracking immediately for the satellite because of the location of the ground station, so we had to wait more than two hours to receive precise information. But we knew right away from the rocket’s data that there was an issue.


What did you do once you had confirmation from the satellite that it was in a bad place?

The satellite triggered an automatic sequence putting it into safe mode, in which it partially opens its solar arrays to generate power. We knew we had to act quickly because an orbit like the one where we were is not a good place to be for a long time. So we had several hours during which to evaluate the risks, inform our customer of the situation, and then make a decision.


Were decisions being made from the


, where the launch occurred?

We had two teams in place — one at our Cannes satellite operations center, and the other in Fucino, where our partner Telespazio has a satellite tracking and control center.


Early on, did you worry that the satellite was going to be a complete loss?

No, we knew from the moment of separation that we would be able to recover at least some life. We had been able to fill the fuel tanks 95 percent before the satellite was integrated into the launcher. With that fuel level, we would have provided 17 to 18 years of in-orbit service if the injection point had been nominal.

We never seriously thought the mission would be a total loss. But it then becomes a question of consulting with the customer and with the insurance underwriters to be sure we all agreed on a common strategy.


What is the biggest challenge you had to overcome to raise the orbit?

Certainly one of the biggest challenges related to our salvage strategy of raising the perigee first. With a perigee that low, the satellite is flying over the Earth at such a speed that infrared sensors are more or less inoperable. In this case, we were able to rely on the star tracker that we use on our Spacebus geostationary satellite design. That was a clear facilitator in this case.


Most geostationary telecommunications satellites don’t have star trackers on board. Is it fair to say that for Palapa D, the star tracker was the difference between a 10-plus-year mission and something much less favorable?

I don’t want to exaggerate its role and I wouldn’t say that without it, Palapa D would have been lost. We could have used the satellite’s gyros to help in the maneuvers. But it was a precious ally in the maneuvers. Without it, the situation could have been more complicated and we would have used much more fuel.


Once you began to maneuver the satellite, what was the procedure?

We conducted three perigee-boost maneuvers using the onboard thrusters, during which the star tracker was very useful, and then three apogee maneuvers.


Will this experience cause you to drop using conventional Earth sensors since you can rely on the star tracker?

In the space sector you want as many onboard redundancies as you can get given the mass constraints. But yes, we do think we can do without Earth sensors, and ultimately without gyros. For now we are looking at what design modifications may be introduced.


Are there any drawbacks to star trackers that would argue against their more-widespread use on geostationary satellites?

Certainly they have a cost associated to them, but that’s true of all hardware and star tracker prices are coming down with the introduction of new technology. One advantage is that you can put a star tracker just about anywhere on your satellite, whereas Earth sensors obviously need to be on the side of the satellite pointing toward the Earth. Also a star tracker is less sensitive to measuring inaccuracies than gyros and needs less calibration.


What advice would you give to future satellite builders who find themselves in a similar situation?

Above all, I should stress the importance of rapid decision-making in the early hours after in-orbit separation. This is a 24/7 situation and you cannot afford to have someone unavailable.

In this case, the decision-making was very quick, and we had good lines of communication opened between us, our customer, the technical teams, the Fucino operation and the insurers.

Peter B. de Selding was the Paris bureau chief for SpaceNews.