The Meteosat Third Generation industrial team. Credit: Thales Alenia Space
CANNES, France — The six-satellite Meteosat Third Generation (MTG) program — the biggest satellite construction contract ever signed by the European Space Agency (ESA) — is running more than two years late and 12% over budget following issues with its cryogenic coolers and micro-vibrations.
The first series of two imaging and one sounding satellite now are scheduled to be launched starting in late 2021, with the identical series launched around 10 years later.
The contract signed in February 2012 was valued at about 1.5 billion euros in today’s economic conditions and foresaw a first launch in 2018. ESA is responsible 62% of the contract, with the 30-nation Eumetsat meteorological satellite organization picking up the remaining 38%.
To accommodate the delays, ESA signed a separate agreement with the industrial consortium led by Thales Alenia Space of France and Italy and OHB SE of Germany valued at some 180 million euros ($202.6 million).
Perhaps because of is size, the MTG contract competition was arguably the most contentious in ESA’s history, with the German government questioning ESA’s Tender Evaluation Board decision in favor of Thales and OHB over Airbus Defence and Space.
Threatened as perhaps never before, the ESA system held, but only after adding some 100 million euros of work for Airbus to mollify Germany.
At the time of the contract signing, then-ESA Director General Jean-Jacques Dordain warned that any cost overrun would be “catastrophic” because they might call ESA’s judgment into question.
With MTG having just passed its Critical Design Review, its program managers from ESA and Thales Alenia Space discussed the reasons for the overruns and the program’s status.
In separate interviews, ESA’s Paul Blythe and Jean-Michel Reix of Thales Alenia Space agreed on all major points as to what happened. Blythe said he is all but certain that the same issues would have delayed the competing consortium, which would have relied on many of the same contractors. The interview here is with Paul Blythe.
Paul Blythe, Meteosat program manager, European Space Agency. Credit: ESA
When did the MTG contract work begin?
The original contract start was in November 2010, under an Authorization to Proceed. We then went through price conversions after they consolidated their subcontractors, and then signed the full contract 18 months later, in February 2012.
Then it gradually became clear that some systems were causing delays?
Yes. The evidence was accumulating, so we came up with a more stable schedule, which showed would need to add several years to the program because of what was happening. So we concluded with a date of late 2020 or early 2021, so about another two years put on the program.
At this point we have added another six months, to mid-2021, for the flight acceptance review, which means we can launch at the end of 2021.
What is the new budget caused by this delay?
There were two riders added to the contract. The first was effectively to re-baseline the schedule and reset the financial frames to make things a bit more stable. At that point, in 2008 terms, we pumped about 150 million euros [180 million at today’s value] into the industrial team.
Following that was a new element, related to storage. We always knew we were going to do this. We entered negotiations for the storage campaign, and then de-storage and testing before launch. That piece injected another 150-160 million into the program. This had nothing to do with the technology hurdles.
So in 2008 terms we are looking at about a 1.7 billion euro contract. In today’s values and we’re at 2 billion euros.
Is 180 million euros in today’s economic value for 10 years of storage, plus refurbishment and testing, within the usual price range so far as you know?
Yes, it’s pretty normal. I did a similar negotiation for the Metop First Generation, and they are all in the same band.
What were the technical challenges that cause the delay to MTP?
Micro-vibration was one. We had an original configuration that was probably optimistic. It was only when we understood the details of all the potential disturbances on board — the reaction wheels and other mechanisms — that we realized micro vibration was a problem.
It required a lot more analysis and we had to introduce isolators and dampers to be able to minimize the disturbance to the instruments. So that was a big one.
Also, the cryogenic cooling loop was an issue. The detectors are operating at about 55 Kelvin. The first cooler that was built worked very well. The coolers themselves are very high tech and had some teething problems. That resulted in delays in the availability of flight hardware.
The detectors also are pushing the edge. The development of the infrared detectors might have suffered from engineers’ optimism at the start of the program. We know all the elements, but the magnitude of getting it to actual flight hardware is often too optimistic.
ESA sometimes signs a payload development contract before the whole satellite development contract to remove some of the technical risk. Should that have been done here?
That was one of the ideas being talked about when we were designing MTG, that we start the instruments early, a bit like Metop Second Generation did, where they kicked off the instruments before the platform, so that if the instruments were delayed we weren’t going to have the whole team associated with the main platform running alongside.
With MTG that is more difficult because it’s very much an integrated system. It’s not a platform and instruments separately. They are working together. But with hindsight I would say yes we might have done it slightly differently.
Jean-Michel Reix, Meteosat Third Generation program manager, Thales Alenia Space. Credit: Thales Alenia Space
We did do quite a lot of predevelopment. For the coolers we got to what we thought was a TRL 6. So we knew the performance we could get. What we didn’t know is that they weren’t industrialized. The first cooler we got performed beautifully. The next one wasn’t quite as good and the first flight one was even less good.
How do you account for the degradation in performance?
There are certain elements we have on this build where it’s almost like a black art. Traditionally in engineering you try and set tolerance levels so that all the builds are the same. But these coolers are so complex it’s very difficult to do that.
You have to match parts with other parts and it’s only when you build a few that you realize: OK, we have to have this one with slightly bigger gaps than the other one.
We were down to playing with micron-level gaps, where one with a 10-micron gap would work, but a 20-micron gap wouldn’t work, or wasn’t as efficient. So we knew how to do it but we didn’t know how sensitive it was to build standard.
This contract was the subject of protests by the German government, which thought its national champion should have won and went so far as to question ESA’s ability to make an unbiased assessment.
It was a very difficult time.
At the contract signing, then ESA Director-General Jean-Jacques Dordain said: “Any cost overruns would be especially catastrophic. it would open room for critics to say we had selected the wrong team.”
I think we have gotten past that. At the time there were lots of uncertainties. The TEBs [Tender Evaluation Boards] have to take everything into account. The TAS-OHB bid was substantially less costly than the Airbus bid.
But I think we would now be in about the same place regardless of the prime contracting team. A lot of the problems have nothing to do with the consortium. The biggest delays we’ve got are with some of the key subcontractors, not with the prime players.
An MTG model in the Thales Alenia Space clearn room. Credit: Thales Alenia Space
 
You’ve just passed the CDR for the program. A CDR in mid-2019 for a launch in late 2021 seems like a tight schedule.
This is an interpretation of the rules of the ESA system. Formally, a critical design review should have results from the key development hardware. We decided we would wait until we got all the major development tests done — the engineering model was build and performed well, we had done the structural testing. We were confident we could go to the CDR and say we are happy with the design.
We recognized this was coming late, so in 2016, to allow us to start moving toward flight with a sensible schedule, we had an intermediate design checkpoint. It was effectively a pre-CDR. this intermediate point two and one-half years ago was to make sure all the major elements were good. But it wasn’t a formal review.
Now we have just done the formal review and the process has been successful. We did the review, presented it to the board and the board closed it on the day, which doesn’t happen very often.
So it is late, but it was successful and as a result we’re well advanced.
What critical milestones are in front of you that are the trickiest?
All the major building blocks are in relatively good shape as we put them together into a flight instrument, particularly the FCI. A major element which happens in mid-2020 is the FCI completed and then it goes into optical vacuum testing. We have tested the elements, and the detectors on their own and bits of optics and the detectors, but we haven’t put the whole thing together yet.
And that optical vac test is proof that we do have an instrument that does work as the sum of its component parts without any major things we’ve missed. We’ve confident but once we get thought that I will breathe easier. The test includes lots of alignment with the optical ground support equipment to demonstrate performance. It’s very complex, both on the instrument side and on the side of the optical ground support equipment. In its own way it is hugely complicated.
To calibrate an instrument like FCI [Flexible Combined Imager], you have to have the ground support equipment be better than the FCI.
How long does this test last?
We hope to be able to do it in about two months. We’ll see. Thales Alenia Space has some experience here with the Meteosat Second Generation.
So there’s no suspicion that a lowball contract bid was made knowing that more would be needed later?
With all these bids you have a commercial interest involved and everybody is always looking on the optimistic side. But the coolers would have probably come from the same place in the competing consortium and they likely would have had exactly the same problem.
And every time we hit the problem, the industrial team was very professional in trying to solve it. I have no complaints with the way they have responded.
And we’re lucky that MSG is a very reliable old satellite. And we hope it stays tat way. We have all four MSGs working, even though the first one is getting old. We have time to put MTG together in a sensible way rather than rushing to get it up.