As the world’s biggest commercial satellite fleet operator, Intelsat is an active participant in most of the discussions that flow through the satellite telecommunications industry.

Ku- versus Ka-band; the future military appetite for commercial bandwidth; the emerging market for mobile broadband connectivity; the health of the commercial launch and satellite sectors — Intelsat is, given its size, front and center on all these issues.

The Luxembourg- and Washington-based company has registered for a $1.75 billion initial stock offering that may or may not occur this year, and is hesitant to talk financials outside its regular earnings calls.

But Intelsat Chief Technology Officer Thierry Guillemin had lots to say about the industry without mentioning EBITDA. Guillemin spoke with Space News staff writer Peter B. de Selding.


Your Intelsat 23 satellite is next up on an International Launch Services (ILS) Proton rocket, the same rocket whose upper stage failed in August. What is the status of this launch?

I will be taking part, with other ILS customers, in a meeting in Moscow the week of Sept. 3 as part of the failure review that is open to non-Russians. What we are being told now is that we can expect our launch to occur in late October. The launch had been scheduled for June, then August. But this is a fairly quick return to flight.


Any concerns about the Proton?

Interestingly, the last three Proton failures have occurred on Russian federal missions, not on ILS commercial missions. The two previous failures — the ones before August — were attributed to human error. It is easier to come back to business when this is the root cause.


Your Intelsat 19 (IS-19) satellite, built by Space Systems/Loral (SS/L) and launched in May aboard a Sea Launch rocket, suffered damage to one of its solar arrays. Sea Launch and SS/L have disputed the cause. What happened from your perspective?

The last time something similar happened with an SS/L satellite on a Sea Launch was in 2004 and the inquiry never settled on a clear cause. The industry as a whole allowed the failure review to finish without a conclusion. All parties involved would not like that to happen this time.

The IS-19 launch continues to be a difficult investigation. Getting to the root cause seems to be a challenge. It will take more time to reach a conclusion.


Inquiries are not cost-free. How long can it continue?

We are in an industry that does not like technical issues that are not explained. We have not reached the end of it yet.


How much power has IS-19 lost permanently as a result of this?

We lost some power, which will result in a loss of some growth potential, but we will be able to serve all our customers and operate the mobility beams on the satellite. I would prefer to wait until after the next eclipse season, starting in August and ending in October, to make a final assessment of the loss of power.


You have just ordered your first Epic satellite from Boeing, a 702 model that provides high throughput in C- and Ku-band. What makes this satellite different?

Epic is a global overlay for our current network. It will be the first ever offered by a single operator. IS-27 is the last satellite we have under our current program to provide global mobility coverage. It will be launched in 2013. The Epic satellites will overlay this mobility network to provide capacity over the highest-trafficked regions of the world.

IS-29e brings additional capacity over the Atlantic, which is exactly what customers like Panasonic told us they need. Customers will be able to go seamlessly from the global Ku-band coverage to the higher-throughput coverage with the same terminals, within the same network, with the same IP address. It’s totally integrated with what we have today.


So customers’ existing Ku-band antennas won’t need to be modified for the higher-throughput Epic satellites?

No. Customers can take advantage of the higher power to reduce their terminal size, or use their terminals’ uplink power with better receive-performance of the satellite. The key to Epic is freedom of choice. Customers choose their terminal technology and their gateway locations at their own pace. So it’s lower capex and lower cost of ownership for them. We don’t put them in a corner where they have to invest millions to change their equipment.

They are also able to groom capacity that before was spread over a number of satellites on a smaller number of larger spacecraft.

You need to look at the overall ecosystem, and the end-to-end economics, to understand why it makes sense for the business case of customers. In the weeks after announcing Epic, we signed three contracts for $500 million. Each is a leader — Panasonic in in-flight entertainment and aeronautical broadband, Harris in energy industry communications, and MTN for cruise ships — in totally different businesses, and these are very sophisticated customers. Government customers have the same reasons to like Epic.


What are the characteristics of the IS-29e satellite?

We are in the 15-kilowatt range for end-of-life power. It has throughput of between 25 and 30 gigabits per second. It covers South and North America and the North Atlantic.

IS-33e, which we will order in a couple of months, covers the rest of the world’s land mass plus high-traffic mobility routes. So the two satellites will offer pretty much global coverage. IS-29e will be delivered in July 2015, with IS-33e to be ordered in the next couple of months for delivery at the end of 2015. IS-29e has a mixed Ku- and C-band payload. We have a split between wide beams and frequency-reusing spot beams in both frequencies, but especially in Ku-band.

Mobile operators want to combine broadband access with video, and we can use this combination of wide beams and spot beams to provide this service.


Your global mobility network includes 10 beams on seven satellites. How do you allocate power between mobile and fixed customers?

This has been three years in the making. It’s like a jigsaw puzzle of Ku-band coverage. The beams are extremely precise. Capacity allocation was determined when we designed the satellites. We have some flexibility in our design, and the ability to steer a Ku beam differently to allocate capacity to other purposes down the road. Both these two satellites were in our capex guidance.


No Ka-band?

Frequency choice is a consequence of your business plan, not a driver of it. What drives the decision? The size of the geographic area is one.

Second, what customers and what applications? Even if there are ways to mitigate the effect of atmospheric losses, the fact is that higher frequencies lose more energy in the clouds. You cannot do anything about that.

Third, the economics that are so important for our customers: Do I want to use different uplink and downlink frequencies? What is my base of installed terminals and how long do I want to use them?

These things will lead to the best choice of frequencies. Believe me, if we have a business case that leads to the choice of Ka-band, we will do Ka-band. We have absolutely no reason not to.


Don’t limits on frequency reuse in certain areas argue for Ka-band?

Not for us. What we are doing with Epic is to unleash a potential of Ku-band frequencies that was totally underutilized, untapped so far. There are Ku-band slots we have that are used with a minimum frequency-reuse multiple. Epic is opening the wealth of frequencies available in C- and Ku-band.


Boeing has a new satellite design using electric propulsion for both orbit raising and for in-orbit station-keeping that saves weight and thus reduces launch costs. Did you consider it?

Electric propulsion is a technology we know well. We used electric with Galaxy 11 in 1999, with conventional chemical propulsion used for the first portion of orbit-raising, meaning getting above the Van Allen belts.

Our preference is to start using electric propulsion above the Van Allen belts because spacecraft electronics and solar arrays do not like to dwell too long in a harsh radiation environment. I have no doubt that Boeing is addressing this problem.

You have to pay for these additional electric thrusters and the cost is pretty high. But the nine to 12 months it takes for orbit-raising with electric propulsion is something we can live with.


Intelsat helped Sea Launch AG return from bankruptcy, and you signed with Space Exploration Technologies (SpaceX) for the first commercial Falcon Heavy launch. At what point does a hypercompetitive launch-services market undermine the long-term viability of the rocket providers?

We are the first announced commercial customer for Falcon Heavy, and we welcome new entrants and trends like SpaceX. In the past couple of years, satellites generally have been delivered on time, sometimes even ahead of schedule. Then they have to wait months before they can be launched. To me this is an indication that there are still ways to put more capacity into the system. We all work on optimizing schedules. In 2012 it is not acceptable to leave an expensive asset waiting for months for a launch.

The incumbent launch-service providers will have to adapt to the new environment. Whoever was on the Proton launch manifest this year knew that it was not going to be easy to be launched on time. It is important to have enough players, and quality players.


The Intelsat 29e is the fourth of a four-satellite order you made with Boeing. You have said you wanted Boeing back in the commercial market. Are there enough manufacturers in the market now?

While there has been intense reorganization and consolidation among satellite operators in the last 10 years, the incumbent satellite manufacturers were already here 10, 20 years ago. This looks like an extremely stable ecosystem. I have heard these concerns about pressure on the margins, and the fact that it is a challenging business — and it is.

But these companies are all still in operation and running. The only notable change is that Lockheed Martin has not yet really come back to the commercial market, although they would appear to have the same reasons as Boeing to do so.


The world’s biggest satellite buyer, the U.S. Defense Department, for a decade has been on a buying spree. That is now over.Won’t that mean more manufacturers focusing on the commercial market, pressuring profit margins that are already thin? 

We’ll see how the evolution in demand plays out. There are trends for different needs — higher throughput, high-resolution video for unmanned aerial vehicles. We will see how these needs evolve.

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