Venice: Higher premiums seem likely for launch insurance, and there is a call for better satellite quality standards, too.

Traditionally, satellite manufacturers, operators and insurers have been at their point of maximum anxiety at the moment of lift-off. Once a satellite has achieved orbit

successfully, everyone tends to sigh with relief and relax. This categorisation of risk exposure has been reflected in the premiums charged for the "Launch" and "On-orbit" phases of satellite operations. While the latter have been as low as 2% of insured value per year, launch premiums have oscillated wildly between 13% and 30%, according to the recent track record of specific launch vehicles.

This was one of the main thrusts of Benito Pagnanelli's opening overview address at the space insurance conference organised biennially by Assicurazioni Generali in Venice. Pagnanelli is deputy general manager of Generali, which was a pioneer in space insurance and is now the world's leading underwriter in this sector.

Healthy growth in capacity

The "capacity" of the space insurance market to provide risk coverage is showing healthy growth, from $550 million at any one time in 1995 to over $850 million today. European insurers provide 71% of this, with Generali itself contributing $110 million. Gross earned premium volume is currently standing at a similar annual figure, according to charts produced by Pagnanelli. But he noted that insured values are also rising, now exceeding $300 million on average for each satellite and $500 million per launch. On the other hand, following the "Annus Terribilis" of 1994 (when losses approached $800 million, generating premiums exceeding 25% the following year), the launch premium trend is falling back to the 15-17% level.

But while launch insurance remains likely to demand significantly higher premiums for the foreseeable future, there is a growing feeling in the space insurance community that satellite performance and quality standards need closer examination.

Specifically, Pagnanelli drew attention to the fact that during the 1995-96 period, payments for claims relating to satellites alone (ie, not including launcher losses) almost doubled, accounting for 38% of total claims. In the previous two-year period (1993-94), satellite claims represented only 20% of the total.

Some of these satellite-only claims had their origin in causes "external to the satellite". Though it is often impossible to determine exactly the cause of an in-orbit malfunction, solar flares (and the surface charging they can induce on a satellite) have been blamed for some anomalies. In particular, the sudden total loss of AT&T's Telstar 401 in January this year has been ascribed to a "solar event". In addition to revenue losses, this event cost AT&T $234.5 million, this being the amount by which Loral Space & Communications slashed its original $712.5 million bid for the purchase of the entire AT&T Skynet Satellite Services unit. AT&T has been unlucky in recent years, since Telstar 402 became "irrecoverably unstable" and lost telemetry contact shortly after separating from its launcher in September 1994. It has since been replaced by another Lockheed Martin satellite.

"Geomagnetic disturbances" were also blamed for loss of attitude control on both of Telesat Canada's Anik-E satellites, also in 1994. Control was later re-established, though at the life-reducing expense of propellant in the case of Anik-E2. More recently, a solar event is being blamed initially for the 15% power loss experienced on 11 April by the Loral-built Tempo-2 DBS satellite owned by TCI Satellite Entertainment. As a result, PanAmSat has decided to delay the Ariane launch of its PAS-6 satellite by some months, for a critical review of its power system and new-technology gallium arsenide solar arrays. PAS-6 is an SS/Loral FS-1300 satellite similar in design to Tempo-2.

Man-made space debris (defunct satellites and spent rocket stages not safely deorbited, likewise fragments from stages in which residual fuel has exploded) has long been feared as a potential cause of satellite losses. The latest of many symposia on this subject has concluded that there are over 10,000 objects in space that are over 10cm wide (or long) and thus trackable by radar; of these, only 5% are operational satellites. Further, there may be anything between 70,000 and 150,000 untrackable objects down to 1cm in size, still quite large enough to disable a satellite.

So far, only one accident has been definitely attributed to such a cause: the severing of a stabiliser boom on the French military research satellite Cerise in July last year was authoritatively ascribed to contact with part of a ten-year-old Ariane upper stage; both objects were tracked by radar. Cerise being a government-owned satellite, this accident did not result in an insurance claim.

Most debris circulates in low orbits, whereas al most all commercial sat ellites at present occupy geostationary orbits at almost 36,000km altitude. But this situation is now changing as the "Big LEO/MEO" constellations of global mobile voice communications satellites begin to be launched into what must be regarded as much more vulnerable orbits. The first five of the Iridium system of 66 satellites were placed in their 780km orbits as this article was being prepared.

It should be noted the behaviour of the sun at periods of maximum activity also poses a greater threat to LEO satellites that it does to those in GEO. The atmosphere tends to expand when heated by high levels of solar activity, and might exert additional drag on spacecraft in very low orbits. Also, ionospheric disturbances generated by solar surges could affect communications traffic, so impacting revenues earned by LEO voice systems.

Now periodic meteor showers have emerged as a new source of risk. Meteors, mostly of sand-grain size, enter the atmosphere continually; but they also form dense clusters containing many larger bodies, orbiting the sun in long-period orbits that suggest they are fragmented comets. One such recurrent shower, known as the Leonids, will pass close by the Earth in 1998/99, at a time when will be over 90 operational satellites in geostationary orbit. So far, no satellite losses have been definitely attributed to meteoric impact (something which would be hard to prove, given that meteors are invisible until they enter the atmosphere), but there have been a number of unexplained satellite "sudden deaths" over the years.

Quality control concerns

But there is growing concern, as Generali's Pagnanelli noted, that total or partial failures of satellites in orbit may also be attributable to human causes dating from prior to their launch. Recent failures for which no external cause has been blamed include partial power losses on Asiasat-2 (Hong Kong) and Indonesia's Palapa-C1.

There is also the vexed affair of spot-beam failures on the near-twin North American mobile service satellites operated by American Mobile Satellite Corp and Canada's TMI Communications. Both the AMSC and MSAT satellites had HS 601 platforms from Hughes Space and Communications and payloads largely manufactured by Spar Aerospace; the two satellites are intended to provide mutual back-up. AMSC-1 was launched in April 1995, MSAT-1 one year later.

Soon after the AMSC launch, it became apparent that one of its spot beams was not performing correctly, yielding reduced power in part of the coverage area. The operator is reported as saying that this corresponded to a revenue loss of around $135 million, though it is accepted that the system was possibly being incorrectly marketed to users of portable terminals rather than larger vehicle-mounted ones. Eventually, insurers agreed to pay AMSC some $66 million.

More disturbingly, similar problems emerged later on the Canadian MSAT satellite, prompting closer investigation by the insurers. These claimed to have unearthed evidence that Spar had not tested the payload equipment correctly. Late last year, underwriters filed suit against Spar, alleging "intentional" concealment of certain defects. The lawsuit was later withdrawn (or set aside) after Spar agreed to negotiate. The case is still pending.

This particular incident was not discussed specifically at the Venice meeting; Pagnanelli merely referred to "litigation being on the rise between insurers and the insured", largely devoted to interpretation of the language used in policies and contracts.

In this context, Pagnanelli did point out that whereas launch failures generally result in total loss of the satellite(s), on-orbit failures often do not. One is then left with a satellite providing reduced performance or shortened operational life, and the consequent need to quantify this in money terms. This in turn introduces the concept of "salvage" or "residual value", often giving rise to protracted arguments and occasional lawsuits. Pagnanelli said that the common practice of regarding a satellite that had "lost 50% of its operative capacity in terms of lifetime" as a "constructive total loss" was "lop-sided in favour of the insured". He added that in the opinion of technicians "There isn't one satellite that has not had an anomaly of some extent, even if it has not jeopardised in any way the outcome of the mission."

In general, Pagnanelli expressed disquiet about a definite trend towards decreasing reliability of satellites. Some of this might be attributed to "constantly evolving technology" introducing new and largely uncharted risks with each new satellite generation. He also said that an equally valid explanation could lie in the fact that "pre-launch tests and qualification of ... new components have been cut down in order to reduce costs and sharpen competitive edge" (see Interavia No. 607, p. 41).

"Lifetime" insurance

Demand for long-duration or "lifetime" insurance is growing as an increasing number of entrepreneurial satellite systems are financed by banks or other money-lenders rather than by the resources of their operators. Traditionally, satellite operators have had a broad spread of business (often derived largely from non-satellite activities), but now companies are entering the market with assets consisting merely of one or two satellites, financed with borrowed money.

If a financing source is dependent on a continuing revenue stream for the return of its investment, it obviously seeks protection for this investment. In Pagnanelli's view, this should be based on yearly re-evaluations of the premiums "on the basis of the satellite's health status." Embarking on a lifetime policy without regular risk re-appraisal amounts to "gambling", he believes. But such reappraisal demands more technical expertise than many underwriters (or venture capitalists) can muster.