The first documentable mention of the possibility of using a free Ariane-5 launch for Cluster was made at the 41st Meeting of the Agency's Space Science Advisory Committee (SSAC), held on 24-25 October 1985. At that meeting, the Implementation Plan for Horizon 2000, which mentioned various options, was presented and discussed. In particular, there were basically two European alternatives to the ESA-NASA cooperation option, which would have entrusted the launch of both SOHO and Cluster to the United States: one was 'the launch of SOHO/Cluster on an Ariane-5 demonstration flight in 1994', and the other 'the combined launch of SOHO and Cluster on an Ariane-4 in late 1993'.
The SSAC had an in-depth discussion on the two Solar Terrestrial Physics Cornerstone missions, and three possibilities were noted:
The SSAC Recommendation on the Implementation Plan for Space Science: Horizon 2000 did not go into the details of the various launch options, probably because the issue was not yet considered mature.
Two important facts appear from this first document:
The cost issue for the STP Cornerstone was felt so deeply at the time that a special meeting of the SSAC was called on 19 November 1985. The costs resulting from the internal analysis and an industrial study exceeded the ceiling set by the Implementation Plan of 400 MAU (Million Accounting Units) at 1985 economic conditions. Roughly speaking, the overall cost of STP, including two launches, was in excess of 700 MAU. According to the Executive, the Cornerstone could cost 560 MAU, on the assumption of de-scoping both missions and having both launched by NASA. The cost would increase to 650 MAU if both (de- scoped) missions were launched on a single Ariane-4. On the other hand, the costs associated with a free Ariane-5 launch were expected to be of the order of 20 MAU.
The situation was considered to be so difficult that the proposal was made to launch only one of the two missions as the first Cornerstone. If this view had been accepted, SOHO and Cluster would have been put into competition against each other. It was therefore discarded because of the recognised synergy between SOHO and Cluster. In the end, the SSAC advised the Executive to adopt the philosophy of starting on a de-scoped SOHO and a three-spacecraft Cluster mission, whilst being ready to up- scope both missions if other solutions could be found. The SSAC asked the Executive to convene the Science Working Teams of both missions as soon as possible.
The Solar System Working Group (SSWG) met on 9 January 1986 and the SSAC held its 43rd Meeting the following day. The SSAC formulated a recommendation on the selection of the next scientific project, in which it manifested its agreement with the recommendations of the SSWG, 'in particular that an STP Advisory Group (STPAG) be set up as soon as possible, whose principal task would be to work closely with the Executive on ways and means to achieve the necessary cost reductions'.
The terms of reference were broad ('to advise on ways of ensuring that the programme complies with the financial boundary conditions established by the SPC, without diluting the scientific validity of the Cornerstone'), and the Executive was asked to provide equally broad support.
The STPAG was immediately convened under the chairmanship of Prof. D. Southwood (Imperial College London), and held four meetings, on 23 April, on 23-24 June, on 17-18 September and on 3 December 1986, respectively. At the third of these meetings three options were presented, which were the result of the evolution of previous proposals.
In the end, the STPAG strongly recommended to the Director of Scientific Programmes that he investigate all possible means to implement 'Option B', which would afford greater autonomy to ESA, as the STP Cornerstone. According to 'Option B', ESA would provide the SOHO spacecraft and four Cluster space-craft. One of the Cluster spacecraft would have a dual-orbit capability and first perform an equatorial mission. NASA would operate the initial equatorial mission as well as SOHO, and would take the Cluster wide-band link data. Later ESA would operate the four Cluster spacecraft. NASA would launch the space-craft designed to perform the dual-orbit mission and SOHO. The launch of the remaining Cluster spacecraft on an Ariane-5 test flight was being investigated.
The third STPAG Meeting was followed by the 45th SSAC Meeting (in Innsbruck, 8 October 1986). Here the SSAC produced its 'Recommendation on the Implementation of the STP Cornerstone', having heard both the previous reports of the STPAG and the opinion of the SSWG. The so-called 'Option-B' was recommended, according to which ESA would 'develop four identical Cluster spacecraft, one of which will be launched by NASA in 1993 into an equatorial orbit.....ESA expects to launch the other three Cluster spacecraft in 1994 on an Ariane-5 demonstration flight'. The text goes on to say that 'The SSAC unanimously and strongly recommends that the scenario be followed with determination by the Executive and that a confirmation of NASA's commitment be sought with urgency'.
At the fourth and last STPAG Meeting (on 3 December 1986), a report was given of a meeting on 8 October between ESA's Director General at the time (Prof. R. Lüst) and the Director of Scientific Programmes on the subject of Cluster. It was clear that the Director General was keen to maintain the STP Cornerstone, but he expressed some reservations. In particular, he 'expressed concern about the planned use of the Ariane-5 test flight for Cluster, because of the inherent schedule uncertainties.....and the risk of flying on a test launch'. The Director General therefore advised that alternative launchers be considered, 'e.g. explore with NASA an STS launch'.
The STPAG made no further recommendations, as it was satisfied with the SSAC recommendation. Apparently, the scientific advisory structure felt that the Director General did not advise further de-scoping of the STP payload, but rather to find other (free) launchers. As other free launchers proved not to be available, the notion of an Ariane-5 test flight was progressively frozen.
The SPC discussed the SSAC recommendations at its 43rd Meeting on 17-18 November 1986 and confirmed as the preferred approach the 'launch of remaining Cluster spacecraft planned for October 1994 by the second demonstration flight of Ariane-5 (assumed to be without charge to the STP budget)'.
The decision to fly on an Ariane-5 test flight was based on the need to preserve the mission by avoiding further de-scoping. It was a rational and unanimous recommendation by ESA's scientific advisory structure. At that time, a single Ariane-4 could not put the four Cluster satellites into orbit.
The political importance of Ariane-5 must have been felt to be a reassuring element. On the other hand, the fact that space is a risky business was well-known: by October 1986 there had been 18 Ariane launches, 4 of them unsuccessful; the Shuttle disaster was still a recent memory. We might add that even the Ariane-4's excellent launch record to date equates to a reliability of about 96% (4 failures out of 64 launches).
There were other competitors hoping to be launched on an Ariane-5 test flight, from both within and outside the Agency. In the following years, the SPC would express its worry that other customers might secure the free flight in place of Cluster.
The first action was the release of the Announcement of Opportunity (AO) for the payload on 1 March 1987. Consistent with the above decisions, the AO stated that the launch vehicle would be an Ariane-5 (flight no. 2). The Principal Investigators, in submitting their responses to the AO, accepted this choice and launch on an Ariane-5 test flight was never questioned during the Science Working Team meetings (the SWT included a number of scientists who participated in the original decision).
By the time of the 48th SPC Meeting in July 1988, there was again 'increasing concern over potential financial-resource needs'. In the meantime, a further Programme Review had been conducted by a small ad-hoc review team, chaired by Prof. H. Balsiger (Univ. of Bern), which made a number of recommendations.
The Balsiger Committee had held two meetings, on 13-14 January and 15 February 1988. As a cost-reducing measure, it examined the possibility of having NASA launch Cluster (as the Director General had originally suggested). However, no satisfactory technical solution was found. A final decision on the launcher needed to be taken by the end of 1988 at the latest.
However, already at the time of the 48th SPC Meeting, it was generally accepted that the Cluster mission would consist of four spinning spacecraft, ready to be launched by Ariane-5 in December 1995. The Ariane-5 opportunity, however, implied some 'non- optimum injection strategy for the Cluster mission, which had to be compensated for within the spacecraft design'. Furthermore, some charges, partly known and partly as yet undefined in 1988, would be levied.
Clearly, the advantage of having a free launch was considered to far outweigh the disadvantages listed above, in the eyes of the advisory structure and of the SPC itself which, at its 48th Meeting, approved the course of action proposed by the Executive, including the Ariane-5 launch, without further mention of the risk connected with such a test flight.
No doubts regarding Ariane-5 were raised at subsequent SPC meetings. At the 49th Meeting in August 1988, the Executive noted that 'it was still planned to launch Cluster on the first or second Ariane-5 demonstration flight, while SOHO remained on the US Shuttle manifest, although SOHO's compatibility with a conventional launcher would be maintained for as long as this was not ruled out by financial considerations'.
The minutes of the 49th SSAC Meeting on 7 November 1988 mention a 'new cost at completion of 482 MAU at 1984 prices, compared with the 460 MAU envelope allocated by the SPC. The Balsiger Committee had reduced the cost and also the risk of subsequent cost escalation. Since the figure of 482 MAU was within 5% of the goal, the Director General had accepted that the project should continue, but that it should be aimed to reduce the cost, bringing it back to the 460 MAU figure'. Furthermore, 'the SPC had taken note and concurred with the course of action proposed by the Executive'.
A Call for Experiments to fly on the Ariane-5 development launch in 1995 had been issued in early 1990, and the intention was to select the passengers by end-1990/early-1991. Ten proposals had been received. The Call created concern among the SPC Delegations, who feared that other spacecraft might be launched free rather than Cluster. However, at its 54th Meeting on 31 May 1990, the SPC was informed that 'the Director of Space Transportation Systems had given assurance that the interests of the Science Programme would be safeguarded when it came to selecting APEX passengers'.
At the same meeting, the Executive informed the Delegations that, 'although Ariane flight 501 was scheduled for April 1995 and flight 502 for October 1995, those responsible for the Ariane-5 Programme were fully aware that Cluster would not be ready for launch until December 1995, and were now working to that timescale'.
In the meantime, work was continuing with Ariane to define the Cluster launch-vehicle interfaces.
A discussion about the possible insurance of scientific satellites had taken place at the 53rd SPC Meeting (20 21 November 1989) in the light of the failure of the Apogee Boost Motor on Hipparcos. According to the Executive, insurance should be discussed on a case-by-case basis. Self-insurance was out of the question, because it simply meant that the ESA Member States would have to commit to provide the money for a replacement satellite. NASA's scientific satellites, it was noted, were not insured. The matter was left for further consideration. However, the almost miraculous recovery of the Hipparcos mission truncated all further discussion of insurance at that point.
At the 56th SPC Meeting (12-13 June 1990), Delegations were 'advised by the Executive that Ariane-5 Project Managers were fully aware of the requirements for Cluster. An APEX flight was still being negotiated'.
At the 60th SPC Meeting (12-13 June 1991), the Executive confirmed that 'agreement in principle had been reached on the launch configuration for Cluster, using Ariane-5. It explained to Delegations that its preference for Cluster, which would be ready for launch as from December 1995, would be to use Flight 501, in case the flight were delayed, which would make the schedules compatible'.
Delegations did not express any preference for either Ariane- 501 or Ariane-502. The important issue was still to make sure that Cluster had a free launch, compatible with its development schedule.
A report was presented at the 61st SPC Meeting (6-7 November 1991) on the situation with regard to the negotiations to fly Cluster on APEX. At that point, Cluster was still on Ariane-502. It would be launched together with Amsat, which had just been selected for the launch.
A further report was presented at the 64th SPC Meeting (4-5 June 1992), according to which Cluster was still on Ariane-502, scheduled for April 1996. A few months later, at the 65th SPC Meeting (3-4 November 1992), a new report stated that 'A detailed review of the APEX programme has recently taken place to consolidate the launch configurations for V501 and V502 in view of the reduced Ariane support programme. Both launches will now use the short fairing and Speltra, and will have a total capacity of 5200 kg. Cluster is currently compatible with both launch configurations. During the review, APEX confirmed that the Cluster schedule was compatible with the V501 schedule'. The document carries the date of 16 October 1992, but a major decision was being taken precisely at that time.
Following an APEX/Artemis/Cluster meeting at ESTEC on 8 October 1992, a passenger review report was issued on 21 October 1992 recommending Cluster's flight on Ariane-501. The main reason given by the review was that 'Cluster will be ready before Artemis and matches the Ariane-501 date more closely'. Besides, putting Artemis on Ariane-502 would 'offer additional time to find a co-passenger'.
On 11-12 December 1992, ESA's Directors of Science and Space Transportation agreed to the above recommendation, to 'assign Cluster to Ariane-501'.
The Ariane Programme Board had been officially informed of the re-assignment in a document, with the following words: 'A review of the requirements and constraints of Cluster and Artemis, and their development timetables, has resulted in the following re- assignment of passengers:
Thus the shift to flight 501 was driven mainly by schedule and flight-opportunity considerations. No specific announcement was made to the SPC, although the shift had long been in the air. The situation was subsequently regularly reported in the SPC documents, but no special mention was made it at SPC meetings.
The SSAC and Cluster Science Working Team were also informed of the final assignment of Cluster to Ariane-501. Neither the Principal Investigators involved, nor the advisory structure, ever objected. In fact, some PI's were happier with a launch on the first flight, because they felt that an enormous effort would go into making it a success. It was felt that the chances of a failure occurring on either the first or the second flight of an Ariane-5 rocket were even, as the second flight of Ariane-1 and the first flight of Ariane-2 had failed. Moreover, the fact that Ariane-5 had been designed for manned spaceflight seemed to offer an even higher probability of success rather than using another launcher, even one with a low failure rate.
It is noteworthy that NASA launched the first Space Shuttle in April 1980 with astronauts on board, and ISAS of Japan still plans to launch its prestigious MUSES-C mission on the maiden flight of their new M-V launcher in September 1996, despite the Ariane-5 experience.
All in all then, there was nothing so extra-ordinary about the ESA decision to fly the Cluster mission on Ariane-501, especially given the great emphasis on cost savings to which the mission was subjected throughout its development. Of course, with the luxury of twenty-twenty hindsight, we all might have taken different decisions along the way.
The above review of the processes and considerations affecting the decisions at the times that they had to be taken, shows that the decision to fly on the first Ariane-5 flight was logically well founded as well as financially attractive.
ESA Bulletin Nr. 87.