Intelsat Chief Executive Stephen Spengler. Credit: Intelsat
KIEV, Ukraine — How did a $400 million satellite just three years into an expected 15-plus years of operations develop a sudden propellant leak and then lose all communications within 48 hours?
The Intelsat IS-29e satellite was declared a total loss on April 18 following the two separate events, both apparently related to its propulsion system.
Intelsat’s most immediate is how best to relocate IS-29e customers, preferably to other Intelsat satellites but if not then to competitors’ spacecraft. That process began almost immediately after the first incident on April 7. As of April 19, the company had placed more than half the IS-29e customers elsewhere.
An unspecified number are still in the process of being placed on other satellites.
Intelsat then will have to decide how to reconfigure its fleet to do without IS-29e while it mulls whether to order a replacement.
IS-29e showed a major fuel leak on April 7. The leak had the effect of a thruster and forced the satellite off its axis and into a slow tumble. Intelsat and satellite prime contractor Boeing thought they were recovering the satellite’s functionality when the April 9 incident delivered the coup de gras, this time taking down the satellite’s telemetry.
Nine days of effort to regain communications with the tumbling satellite were of no use. Intelsat declared IS-29e, which was not insured, a total loss on April 18.
IS-29e now joins several hundred other dead satellites that will travel the geostationary-orbit arc along the equator for millennia.
With a predictable orbit and well-tracked by ground sensors, the satellite will not be hard to avoid for other geostationary-satellite operators. As IS-29e migrates into and out of other orbital slots, operators with satellites there will perform slight maneuvers to keep out of its way. They are used to this kind of thing.
But IS-29e is one more piece of uncontrollable debris whose batteries have not been drained and whose fuel tanks have not been depressurized, creating a future risk of explosion and debris proliferation along a major orbital highway.
One of the more surprising aspects of the IS-29e failure is how little is known about its cause. This was a high-end spacecraft — a Boeing 702MP model — equipped with the latest telemetry capability.
Intelsat Chief Executive Stephen Spengler discussed what Intelsat knows and doesn’t yet know about what happened with IS-29e. He expressed hope that the detailed investigation by Intelsat and Boeing Satellite Systems International will be able to provide answers.
IS-29e developed what was described as a fuel leak on April 7. What was the first sign of trouble?
The telemetry we were able to gather at the time gave an indication that it was leaking propellant. The telemetry is pretty clear at that point in time. What we don’t know is how it happened, why it happened.
It typically takes quite a bit of time to get through these failure analyses and reviews, a few weeks or months. It starts with a very wide range of root causes and then they go through it.
There is speculation all across the board on what the cause could be and none is weighted more strongly than another at this point.
Nothing occurred prior to the leak that might have caused it so far as you know? It came out of the blue?
Correct. Subject to looking at the telemetry in more detail, there was no event that we saw prior to it. It was just normal operations. But there was obviously a cause. It’s not like these things just rupture, also I guess it could be possible.
Then the propellant leak puts the satellite into a spin and we had to recover.
So you are not ready to point a finger at a piece of debris or other foreign object?
No. It’s too early. We haven’t gotten that far into it.
What happened next?
When the original event occurred, the payload powered down immediately. We weren’t transmitting and customers were off immediately. That’s the way the system is designed.
We didn’t lose command and control of the satellite at that point. We were able to start to work it back into Earth pointing after that initial event. Then on April 9, when we had actually brought it back to Earth pointing, the second event happened.
What were the characteristics of the second event?
Similar to the first: something around the propellant in the fuel system that caused the second anomaly to occur. But with the second event, we weren’t getting the same telemetry from the satellite.  We were getting some, but not much.
With Boeing’s advice and consultation — our engineers and theirs were working this issue around the clock — we began a series of commands to the satellite that we continued to send until yesterday [April 18].
But the telemetry signals from the satellite actually faded away pretty quickly, within a couple of days after April 9.
We had a series of commands that were recommended by Boeing, which we executed on, and it didn’t respond. So the conclusion was: We’re done. We’ve done everything we could possibly do, and the right decision was to declare it as failed.
It doesn’t mean that we wouldn’t try to send more commands, but we don’t have much expectation that we’re going to be able to communicate with the satellite, since we’ve had no word from it since shortly after April 9.
What maneuver were you conducting on the 9th? What was its goal?
On the 9th we had it back to Earth pointing, so we were basically looking to commence restoration of the payload and customer services when the second anomaly occurred.
And this anomaly had the same profile, it looked like some sort of release or leak in the propellant system. But then we lost telemetry.
The leak from the 7th had continued to the 9th, correct?
Yes, most likely.
Could the second event just be a consequence of the first?
It could be, but the problem is we don’t know. We just know these symptoms occurred at these two different times. And now we really have to figure out everything: What caused the first one, what caused the second one, what happened in between — those are all things that will be determined in the failure review process.
Would you say that Boeing has acted as you would hope and expect them to act as prime contractor since the April 7 problem?
Absolutely. Boeing’s engineers were on it right from the moment we contacted them on the 7th, and they were with us throughout the whole process.
With the SES AMC-9 satellite failure in 2017 there were indications of a debris field and a debate over whether the debris was leaked fuel that had frozen. Is there any indication of debris around the satellite?
It’s too early to understand what exactly occurred. There is this video that is out there — http://bit.ly/2ULXKlg —and that people have seen. I have been telling our people: Don’t look at that video with an untrained eye, because you may not be able to interpret it.
ExoAnalytic Solutions produced this video of the Intelsat-29e anomaly. Credit: ExoAnalytic Solutions
Second, I understand that video is not real time. It is accelerated so it wouldn’t necessarily appear as it did in real time. But there could be any number of reasons for why it looks the way it does on video. That’s one fo the things we are going to have to find out from various sources — independent commercial sources that are out there, as well as the Air Force and other places that may be monitoring space activity.
So no indication of any debris or what that might be around the satellite?
No, it has not yet been determined. But at this point I would expect to have been informed by CSpOC [U.S. Air Force Combined Space Operations Center] if there were more than one object there, and we haven’t.
At what point did you tell other operators: We have a satellite drifting out if its box and we can’t control it?
By our license and regulation we informed the FCC [U.S. Federal Communications Commission] immediately that the satellite was going to be drifting out of the box. CSpOC was notified right off the bat as well. They do all the coordination with other operators when the satellite drifts, so it’s normal procedure.
A big fleet operator like Intelsat has the advantage of being able to move customers around to other satellites in its fleet. How has that gone?
Any time you have a situation like this, how quickly a customer can move to another satellite on our fleet or on another satellite fleet will depend on the particular customer’s network.
Mobility customers, for example, that have three-axis-stabilized antennas that can point to any satellite, pretty quickly can have their services recovered — faster than, say, a fixed VSAT customer that has hundreds of sites on a particular satellite that will have to be re-pointed.
The fleet of Epic HTS (high-throughput) satellites is Intelsat’s most-valuable asset, designed to carry the company into a data-centric future. Credit: Intelsat
We engaged with our customers right away when this happened and our teams got engaged and working on restoration plans for our customers across the board. At this point most of our customers have been restored. A lot of that has been on the Intelsat fleet.
But we have had the support of other satellite operators as well. That has been really appreciated, how our colleagues in the industry have responded has been excellent with respect to our customers.
It’s not pure charity if they’re getting new customers….
Well, here’s how it works. In our particular case we already had an established restoration agreement, reciprocal, with one of our main competitors and that has worked well.
It’s not that this happens all that often, but we have experienced in the past that most companies step up and work with us, and recognize that this is is tough time. Of course there is gain to business for them in those arrangements. They are not doing it for free, they are getting compensated for it as well.
So you would say their reaction hasn’t been just mercantile, where they’re looking to profit as much as they can?
That’s correct, it was fairly negotiated.
Is your IS-32e, with an HTS payload and not far away from the IS-29e slot, available to you here for IS-29e customers?
It is available, but it is largely full and pretty heavily utilized. But it was helpful for our Flex managed service. We did have some capacity there to support our customers.
I would think that might work for a mobility customer like Panasonic Avionics.
With Panasonic we have had to work some other options to provide them with some restoration options. IS-32e would be suitable for them if it were available. I’d rather not get into specific customers. We have restored over 50% of them.
Wouldn’t customers avail themselves of the alternatives immediately?
Sometimes it takes time to reconfigure a network on the ground, and repointing customer antennas. In some cases they say: Great, we can take it right now. In other cases they have some engineering work to do.
So you advise customers of a failure and then, if they cannot be placed on your fleet, you advise them o other options and suggest they go talk to another operator?
We work with the customer with the solution for the customer, whether it’s on an Intelsat or a third-party satellite. We are with them throughout the process.
What is the split so far between customers being moved onto other satellites in your network, and those being relocated to other operators?
I don’t want to give that split right now. But having a fleet as large as ours, with 54 satellites, there are other options. We had some capacity available on another Intelsat Epic satellite, Intelsat 37e, that we have been able  put our customers on. But you dot always have a like-for-like, high-throughput spot beam to offer everyone. So we had to work different kinds of arrangements.
You have elected not to purchase in-orbit insurance beyond the fist year after launch. In-orbit insurance is inexpensive now. Is this something you might reconsider?
This has been our practice for awhile. We typically do not have in orbit insurance except in certain circumstances. For the most part we self-insure those operations past the first year. We have a large fleet and when those things have happened we’ve been able to restore most of the services. 
It depends on the size of the anomaly. I am going hack in my 16 years here as a reference point. So we just haven’t done it, and most of the issues that we have encountered, in terms of the need for insurance, they typically happen during launch, or post-launch early orbit raising through the early days of operations. This is really an unusual event for a satellite of this age.
What is the estimated revenue impact of this failure based on 2018 revenue from IS-29e?
We’re not going to take about the financials right now.
Is the loss of IS-29e going to trigger the purchase of a new satellite or is that not clear yet?
We are thinking about it but we haven’t spent a lot of time on those going-forward plans yet and how we would replace this capacity. There are certain things we could do within our existing fleet