Commercial Communications Satellite Bus Reliability Analysis

August 2004
(courtesy of Frost & Sullivan)

Executive Summary

Frost & Sullivan has analyzed the on-orbit performance of the major commercially available satellite buses and considered the strengths and weaknesses of their manufacturers in order to determine which satellite bus (or platform) is more reliable. Based on both Frost & Sullivan and Airclaims data, this study highlights reliability records, anomaly trends, and the impact of these factors on the insurance industry and hence, the satellite industry overall. It presents the results of Frost & Sullivan’s research and further discusses the state of the space insurance industry and its affect on the cost of satellite ownership, a particularly important issue in this time of slowing satellite industry growth.
In recent years, satellite manufacturers have been criticized for an increase in the rate of inorbit anomalies and, beyond that, complete failures of commercial communications satellites. In terms of satellite insurance claims, the period from 1998 through 2001 was particularly bad. The unusually high number of satellite anomalies and resulting insurance claims have seriously affected both the quality and reliability of services provided by commercial satellite operators and have (along with notable launch vehicle failures) had a negative impact on investors’ perceptions of the space industry as a whole. Beyond that, such problems have resulted in billions of dollars of losses for space insurance underwriters, increasing space insurance premium rates and hence the cost of ownership for commercial communications satellites in general.
Although the last two years have seen a reduction in the number of serious anomalies the affects of the 1998-2001 period remain. Insurance costs have risen considerably and attitudes towards satellites and their manufacturers have changed. Before 1998 the satellite industry and its customers were moving toward a vision of satellites as a commodity. Satellites were expected to function well and new technologies to expand their capabilities were embraced. Satellite manufacturers built new manufacturing facilities and anticipated ever-increasing orders. This vision proved faulty when the new technologies showed flaws once in service and previously reliable satellites began to develop problems as well. The large market for satellites that had motivated the more production-orientated manufacturing techniques failed to appear and the commodity model of satellite manufacturing has now generally been abandoned.
In response to these ongoing events, Frost & Sullivan recommends that satellite operators focus even more attention on reliability and make more use of the total cost of satellite ownership as a decision making tool. Operators should factor insurance costs into their procurement decisions.
To do this, it is important that they choose proven, reliable satellite buses that will support a low cost of ownership over the life of the spacecraft, while providing the quality of service and operations necessary for successful competition in the telecommunications market. This study is intended to provide information useful in such decisions.
Satellite manufacturers need to pay close attention to manufacturing quality and customer needs. A race to the bottom in satellite pricing will not solve the industry’s problems. Rather satellite manufacturers need to engage in competition based on the total cost of ownership of their products. Only by providing truly superior solutions to customer needs will manufacturers be able to expand their market share in these lean times. Frost and Sullivan hopes that this study will aid in the development of a value, not price, based satellite industry.
Based on careful analysis of the data, Frost & Sullivan’s key findings are that of the buses currently available, the Lockheed Martin Commercial Space Systems (LMCSS) A2100 and the Boeing Satellite Systems (BSS) BS 376 have the lowest number of satellites with claims, as a percentage of those satellites of each bus type launched. Unlike the small Boeing 376, however, the A2100 bus covers a broad range of satellite applications and power levels. For this reason, according to the study criteria, Frost & Sullivan considers the A2100 to be the most reliable satellite now available for a majority of satellite applications. In addition, the EADS Astrium Eurostar 2000 bus has achieved a very low value of insurance claims. Although relatively more claims have been made against the Eurostar, they have been for small amounts.

return to top

Introduction

Since 1998, the satellite industry has suffered from a series of major systemic problems involving a number of different satellite buses and manufacturers, most notably the Boeing 601 and 702, which suffered from systemic problems with solar arrays (702), on-board processors (601), and ion propulsion subsystem (601). These failures have created an operational and financial crisis in the satellite industry. Operationally, the repeated loss of satellite capacity (in shortened life spans and the partial or total loss of on-orbit transponder assets) has caused immediate problems for satellite operators as in the loss of Galaxy 4 in May 1998, which affected millions of pagers across the United States. In broader view, problems such as those on XM Satellite Radio’s Boeing 702 satellites have damaged business plans and threatened some operator’s financial survival. Operators have not only lost income in the immediate aftermath of such incidents, but have also seen expenses, in the form of satellite insurance rates, rise precipitously, noticeably increasing the cost of replacement satellites. In this environment, satellite reliability has become an issue of paramount importance. In this study, Frost & Sullivan assesses the reliability of the various commercially available communications satellite buses and considers the nature of the recent failures and their affects on the space insurance industry.

return to top

Methodology

For this study, satellite reliability is measured by the percentage of a bus family’s on-orbit population that suffers from at least one serious anomaly (an anomaly affecting the satellite’s ability to generate revenue). In performing this calculation, Frost & Sullivan counted only those anomalies that have resulted in insurance claims, in order to avoid problems of biased and incomplete data as well as issues of data comparability.
An alternate method would be to examine total bus anomalies, but this approach fails in the face of the difficulty of obtaining highly proprietary satellite anomaly data. Without a common source of unbiased data, any analysis would risk severe distortion of its results, through the use of incomplete data (possibly unfairly favoring one manufacturer over another) and would risk the use of non-comparable data (as the result of manufacturer’s different methods of tracking and classifying anomalies). Insurance claims, on the other hand, are publicly available information available from a single reputable source.
This study is based on claim data has been drawn from the Airclaims SpaceTrak TM database, supplemented by additional data from Frost & Sullivan’s proprietary databases. The Airclaims database tracks significant events in the life of commercial satellites from construction to retirement. Airclaims data is an industry standard, widely used by a variety of groups including insurers, operators, manufacturers, and financial institutions.
Airclaims data is used to provide consistency and validation of the Frost & Sullivan data. By using both databases, Frost & Sullivan ensured that the information in this analysis is precise and unbiased. In addition to using the SpaceTrak data, Frost & Sullivan worked with Airclaims’ Space Analyst, David Todd, in the overall design of the research and obtained Mr. Todd’s advice on interpreting data and in preparation of the report (all calculations are the responsibility of Frost & Sullivan).
The satellite buses evaluated in this study are those that have accumulated a considerable amount of on-orbit experience with commercial operators and are readily available on the commercial market. The Alenia Aerospace Italsat bus and Orbital Sciences’ Starbus are excluded from this report because of their limited on-orbit records. Other buses, in particular those from Russia, China, and India, are not truly part of the commercial marketplace. If any of them were to be widely accepted by global commercial operators, they would merit consideration in a future study. The remaining commercial buses are evaluated over their total commercial experience through March 2004. Non-commercial satellites based on these buses are not considered (because of differences in insurance coverage, systems, and operation), but all of the commercially procured, commercially operated, variants are considered. Of commercial satellites that fit the criteria above, only those satellites that were involved in launch failures are excluded from this study.

return to top

Key Findings

A majority of the claims made against the satellites considered in this study were bus, rather than payload, related. Of the major commercial satellite buses currently available and in widespread use, the Lockheed Martin Commercial Space Systems (LMCSS) A2100 and the Boeing Satellite Systems 376 have the lowest rate of buses with anomalies (resulting in claims filed) as a percentage of satellites put into service. The small Boeing 376, is no longer well positioned to fill the majority of operator needs, however, the A2100 offers models that cover a broad range of satellite uses (see Figure 3). For this reason Frost & Sullivan considers the A2100 to be the most reliable satellite now available for a majority of satellite applications according to the criteria used in this study . In addition to these satellite buses, EADS Astrium’s Eurostar 2000 receives a special commendation for the very low reported claim values resulting from anomalies on this bus (see Figure 5).

return to top

Overview of Manufacturers and Buses Studied

There are currently five companies providing satellites that cover the range of sizes required by the commercial communications satellite market. From the United States come Boeing Satellite Systems (BSS), Lockheed Martin Commercial Space Systems (LMCSS), and Space Systems/Loral (SS/L). The remaining two are the Europeans Alcatel and EADS Astrium. Figure 1 shows maximum EOL Spacecraft power for actual satellites (not projected maximums).

Figure 1: Major Commercial Satellite Buses

Satellite Bus

Manufacturer

Maximum EOL Power

Minimum EOL Power

Successful Commercial Deployments*

A 2100

Lockheed Martin Commercial Space Systems

13100

2450

23

BSS 376

Boeing Satellite Systems

990

990

55

BSS 601

Boeing Satellite Systems

8800

3300

62

BSS 702

Boeing Satellite Systems

17600

7200

8

Eurostar 2000

EADS Astrium

8000

3000

19

LS 1300

Space Systems/Loral

16000

4500

41

Spacebus 3000

Alcatel

9000

5000

18

* Through March 2004, satellites lost in vehicle-related launch failures not included

Source: Frost & Sullivan

return to top

Satellite Manufacturer Profiles

Boeing Satellite Systems
Boeing Satellite Systems (BSS) is the result of the Boeing Company’s January 2000 purchase of Hughes Electronics’ Space and Communications division. Hughes Space and Communications built the world’s first geosynchronous satellite and had always been a leader in the commercial satellite industry. The HS-376 (now BS 376) series is still a watchword for satellite reliability. Things have not gone well, however, for the renamed Boeing Satellite Systems. Organizationally, Hughes turned out to be a poor match for Boeing’s corporate culture and BSS has suffered from quality problems. Although some observers see a relationship between this and the quality lapses suffered by the BS 601 and BS 702, incidents of all of the systemic BS 601 and BS 702 failures occurred on satellites that were built by Hughes before the Boeing takeover. As other, Boeing-built, satellites suffer from the same problems, it is difficult to see the change in cultures as the root cause for the systemic problems. Another argument is that the very success of Hughes’ spin stabilized 376 slowed the development of three-axis stabilized satellites such as the 601 and 702 and that relatively late start contributed to some of the 601 and 702’s technical issues.
As a result of these problems, BSS has had an even more difficult time getting orders than other manufacturers (though no one has done well in the depressed industry). Events reached their nadir when Harry Stonecipher returned to Boeing as President and CEO with the expectation of closing BSS. A turnaround plan has since been developed and Boeing expects a return to profitability for BSS in 2005. This plan is predicated, however, on a mix of two-thirds government business and only one-third commercial. In the future, BSS will bid on only those procurements in which it feels it will have a special advantage. Another recent Boeing initiative is an agreement to cooperate with the Indian Space Research Organization on Indian built communications satellites. It would appear that Boeing is still concerned about the commoditization of the satellite market and is planning to support its operations on government business while cherry-picking select commercial opportunities. BSS will be a continuing player in the commercial satellite business, but even if these plans are realized, it is unlikely to be an overall industry leader.
Lockheed Martin Commercial Space Systems
Lockheed Martin Commercial Space Systems (LMCSS) is the result of a series of mergers over the past 18 years. Its heritage includes RCA Astro Electronics, GE Space Systems Division, GE Astro Space/ Martin Marietta Astro Space and Lockheed’s own satellite manufacturing operations. Like Boeing, LMCSS is part of a major defense contractor, a fact which has fueled concern that Lockheed Martin would leave the commercial market as Boeing nearly did. In fact, in 2003 Lockheed announced its decision in its annual report to build on its core heritage as an aerospace manufacturer and continue in the commercial space industry. Where a few years ago Lockheed Martin was expanding into satellite operator’s territory with acquisitions such as Comsat and partnerships like that with Intersputnik, it has since forsaken its quest to become a major satellite operator and focused its efforts on manufacturing. Such a focus on manufacturing and technology (always a Lockheed Martin strong point) speaks well for the future of its satellite manufacturing division, which is well prepared with its A2100 bus. In addition, Lockheed Martin has invested in state of the art manufacturing facilities, which position LMCSS well to compete for the available satellite procurements. Free of the sort of baggage that almost brought its rival BSS down, and part of the larger Lockheed Martin enterprise (now refocused on technology and manufacturing), LMCSS should have a strong future as can be seen by its growing order book.
Space Systems Loral
Space Systems Loral (SS/L) is another satellite manufacturer who’s future has been called into question by concerns about its parent company. SS/L’s parent company Loral Space & Communications filed for Chapter 11 bankruptcy protection in July 2003 and with that filing called into question SS/L’s continued existence. For a while it seemed possible that Loral would be sold to another satellite operator (Echostar was frequently mentioned) who would dispose of SS/L. However, the sale of six satellites to Intelsat for just over $1 billion has given Loral, and with it SS/L, a new lease on life. Loral now plans to refocus on its remaining international communications satellites and its SS/L satellite manufacturing business.
As Loral emerges from bankruptcy, SS/L will have the opportunity to regain its place in the satellite manufacturing industry. SS/L has a strong technological base and has spent the last six years developing its latest generation of satellites, as seen in the latest LS-1300s. It has also recently entered into a partnership with Russia’s Energia to manufacture and market small satellites. Not part of a major defense contractor, SS/L feels that it is better able to fill commercial satellite operators’ needs because this is its single focus (although it has also produced space station components and satellites for government use such as MTSat-1R for Japan). SS/L’s satellites have experienced systemic problems with solar arrays, but if given the opportunity, SS/L should continue to be an able competitor technologically. The real issues for SS/L are whether it can regain sufficient creditability in the marketplace to attract customers and whether it will be able to maintain an attractive cost structure in competition with its larger competitors.
Alcatel Space
Like SS/L, Alcatel has not been part of a major defense contractor (although it acquired Aerospatiale’s manufacturing facility). A product of a series of European mergers, Alcatel is also active in fiber optic cable, wireless infrastructure, enterprise communications solutions, and other network and communication technologies. Most recently, it signed a memo of understanding with Italy’s Finmeccanica to merge their operations. Alcatel will hold two thirds of the new venture, while Finmeccanica will have a one third share, producing the world’s third largest space firm. In addition to its complete satellite business, Alcatel is also known for its satellite communications payloads that have seen use on a wide variety of different platforms including Russia (Express A and AM), China (DFH-4), SS/L (Europe*Star), BSS (XM Radio), and EADS Astrium (Arabsat 4). Although Alcatel recently suffered through the messy breakup of a partnership with SS/L, it is well-positioned to continue in its chosen niches.
While its communications payloads and Spacebus 3000 satellites stand it in good stead, Alcatel is also looking to the future with the Alphabus, which it is developing in cooperation with EADS Astrium. Unlike SS/L, Alcatel is currently profitable (all business segments were positive for the first quarter of 2004) and there is no strong reason for it to leave the satellite manufacturing business at this time. For Alcatel, the largest concern has been that it is first on everyone’s short list for a merger. It is widely (although not universally) felt that Europe would be better served if it had only one satellitemanufacturing arm (EADS Astrium), just as it has only one major launch vehicle supplier (Arianespace).
EADS Astrium
The European Aeronautic Defense and Space Company (EADS) is Europe’s answer to American aerospace giants Lockheed Martin and Boeing. Astrium itself is the result of a 2000 merger of the space division of Daimler Chrysler Aerospace (Germany) with Matra Marconi Space (France, UK) and Computadores Redes e Ingenerio SA (Spain). It became EADS Astrium in June of 2003 when BAE Systems sold EADS its 25 percent share and in January 2004 it added Spain’s CASA Espacio. Like Boeing and Lockheed Martin, EADS is an industry giant and also like them its space division produces spacecraft for a broad range of uses (science, earth observation, and communications). Both financially and technologically, EADS Astrium is in a strong position. Its Eurostar 2000 series has had very low insurance claims and its larger Eurostar 3000 has survived its inaugural launch. If the Eurostar 3000 continues to function well and avoids major development issues (it is after all, a new design and contains new technologies such as Lithium ion batteries) EADS Astrium will be well poised to move forward on a variety of fronts. One of these is clearly commercial communications satellites.

return to top

Industry Overcapacity

Whatever their individual strengths, the same problem faces all commercial satellite manufacturers - insufficient demand for their products. It is difficult to see how five major satellite manufacturers (and a number of smaller ones not dealt with in this study) can survive in a market in which 20 orders a year is seen as an optimistic projection. Prior to its recent merger arrangements, Alcatel was often considered the most likely European to drop by the wayside, with SS/L the obvious candidate to reduce the North American three to two. With Boeing’s turn away from commercial markets and Astrium’s bulking up, it is no longer so easy to make predictions.
There is still a broad perception that, in the long run, Europe’s satellite manufacturers will be merged. This may be true - the Spacebus and Eurostar platforms and their variations do share many characteristics - but there are other considerations. Although Alcatel officials have publicly stated that they still consider some sort of partnership with EADS Astrium a possibility (even after their merger with Finmeccanica), it is not clear that European regulators would allow the loss of competition entailed by such a merger. Likewise it might prove to be more difficult to merge two equals than it would have been for EADS Astrium to swallow the smaller Alcatel. If such a merger was accepted however, it would likely lead to the elimination of the overlapping product lines, along with staff reductions. There are significant duplications in the facilities of Alcatel and Astrium, and the primary activities at each. Concerns have been expressed with the duplication of activities at their Toulouse and Cannes satellite manufacturing centers in particular. Beyond this the role of labor unions cannot be underestimated.

Figure 2: Historic EOL Satellite Power Ranges

Source: Frost & Sullivan

In the United States it seems unlikely that either Boeing or Lockheed Martin would be allowed to buy SS/L for antitrust reasons. It also seems unlikely that the current satellite market would attract a suitor wishing to enter the commercial satellite manufacturing market through a purchase of SS/L. Rumors have mentioned existing aerospace defense players looking to expand into the satellite manufacturing market considering SS/L a possible path to gaining the proper facilities, although SS/L’s fit with a military driven product line is questionable.
Keeping these considerations in mind, there is no clear path to industry consolidation on either side of the Atlantic, although both sides have considerable over capacity and many observers feel that consolidation is in order.

return to top

Historical Satellite Reliability

Of all of the satellites considered in this study, the BSS 376 and the LMCSS A2100 have the best records in terms of the number of satellites with insurance claims. According to Airclaims data, the BSS 376 has had claims made against six of the 55 successfully deployed commercial communications satellites, for a claim rate of 11 percent (see Figure 3). LMCSS has done nearly as well, with a 13 percent record (three claims out of 23 successfully deployed commercial communications satellites). The next best satellite bus is the LS-1300, with a 20 percent claim rate.

Figure 3: Anomalies Resulting in Insurance Claims by Satellite

Rank

Bus

Successful Deployments

Satellites with Insurance Claims

Percentage of Fleet with Claims

1

BSS 376

55

6

11%

2

A2100

23

3

13%

3

LS 1300

41

8

20%

4

BSS 601

62

14

23%

5

Spacebus 3000

18

5

28%

6

Eurostar E2000

19

6

32%

7

BSS 702

8

6

75%

 

Source: Frost & Sullivan and Airclaims

 

although the BS 376 is the most reliable satellite considered in this study, it is also by far the smallest, with EOL power only up to 2 kW. While the BS 376 cannot be ignored (due to its excellent record and large number of deployed satellites), for most modern applications it is no longer truly competitive in size. The LMCSS A2100 is much better placed for the current market, where more than half of demand is in the 6 to 10 kW range and much of the remainder falls in the lower 3 to 6 kW area (still exceeding the available power on the BS 376). As a result, Frost & Sullivan recognizes the LMCSS A2100 as especially notable, as its on-orbit fleet has one of the lowest percentages of insurance claims generating onorbit anomalies (see Figure 3)

Figure 4: Anomalies Resulting in Insurance Claims

Rank

Bus

Successful Deployments

Number of Insurance Claims

1

BSS 376

55

6

2

A2100

23

4

3

LS 1300

41

10

4

BSS 601

62

16

5

Spacebus 3000

18

5

6

Eurostar E2000

19

7

7

BSS 702

8

6

 

Source: Frost & Sullivan and Airclaims

 

Figure 4 shows the absolute number of claims per bus. Because this does not take the number of deployed buses into account, it is less informative than Figure 3, but it is worth noting that of 54 claims only 6 are on a satellite that had a previous claim. Close to 90 percent of satellites that suffered an anomaly did not suffer another, although this may be the result of underwriters’ refusal to insure systems that have had problems.
There is, however, another aspect of insurance claims to consider - claim value. EADS Astrium’s Eurostar 2000 bus is clearly the platform with the lowest value of claims. Although it is never good to have an anomaly occur, the Eurostar 2000 is notable for the low value of the claims against it (see Figure 5).

Figure 5: Value of Insurance Claims

Rank

Bus

Successful Deployments

Value of Insurance Claims (US$M)

1

Eurostar E2000

19

60

2

BSS 376

55

125

3

Spacebus 3000

18

326

4

A2100

23

375

5

LS 1300

41

791

6

BSS 702

8

1040

7

BSS 601

62

1468

 

Source: Frost & Sullivan and Airclaims

 

return to top

Analysis of On-Orbit Anomalies

After examining the causes of the claims considered in this study, Frost & Sullivan has determined that a majority of claims are bus, rather than payload, related. Considered by either number of claims or by total claim values, buses introduce more failure risk than payloads. This conclusion reinforces the importance of bus reliability as a selection factor in satellite procurements and further validates an approach to satellite reliability based on bus type.
Many of the recent satellite reliability problems have resulted from the technologies developed to build larger, more powerful, satellites. A major portion of this development has involved increases in available power and extensions of satellite design life. As a result, it is not surprising that power systems (solar arrays and batteries) and propulsion systems account for 64 percent of the number of insurance claims (see Figure 6). This analysis of claims also shows that power and propulsion systems are the most expensive areas in which anomalies can occur, with 74 percent of the value of all claims (see Figure 7) being generated by 64 percent of the claims.

Figure 6: Number of Insurance Claims by Anomaly Type

Source: Frost & Sullivan and Airclaims

Solar arrays are a particularly important component and serve as a good example of a number of major issues in satellite quality. Fifty percent of the value of insurance claims in this study was the result of solar array problems (38 percent of claims). Because solar cells are often bought rather than built in-house, they are a reminder that prime contractors remain responsible for evaluating the quality of their subcontractors, a practice that some suggested could be left to the subcontractors under the commodity satellite model. Instead, it is clearly necessary for prime contractors to take the lead in maintaining satellite quality. The relative balance between maintaining schedules and assuring quality has shifted in response to the increased number of problems dating from the late 1990s. The increased willingness of both satellite operators and manufacturers to accept reasonable delays in order to assure quality speaks to the experience of the past decade.

Figure 7: Value of Claims by Anomaly Types

Source: Frost & Sullivan and Airclaims

Time is also a factor in satellite failures. As can be seen in Figure 8, over half (64 percent) of the insurance claims that are made occur in the first two years of service. Likewise, 61 percent of claim value accrues in the first two years (Figure 9). By year three, over three quarters of claims (79 percent) have been made and almost three quarters of claim value (74 percent) has been recorded. This failure data came from Airclaims data and reflect an assessment of the claim value as of the date of the actual failure.
Although three quarters of problems overall occur in the first three years of service, there are two types of failures that do not follow this pattern. The first of these is propulsion failures, where half of the failure take place in years one through four but the remaining failures continue until year 10 (See Figure 8). An even more extreme case is central processor failures, which begin in year six, with one each in year 8 and year 9. Clearly both of these systems take longer to reach their failure modes (although the occurrence of propulsion failures starts quickly, with almost 50 percent in the first two years of service). It should be noted, however, that elderly satellites are not usually insured and most are retired (due to exhaustion of fuel) before old age problems really begin to occur.

Figure 8: Annual Anomalies by Type of Claims

Year

Antenna

Battery

Control Processor

Payload Electronics

Propulsion

Solar Array

Structures

Transponder

Total Claims

% of Total Anomalies

1

5/100.0 %

1/17.0%

 

5/71.0%

2/25.0%

8/38.0%

1/100%

2/67.0%

24

44%

2

 

3/50.0%

 

 

2/25.0%

6/29.0%

 

 

11

20%

3

 

2/33.0%

 

2/29.0%

 

4/18.9%

 

 

8

15%

4

 

 

 

 

1/12.5%

1/4.7%

 

 

2

4%

5

 

 

 

 

 

1/4.7%

 

 

1

2%

6

 

 

1/33.3%

 

1/12.5%

 

 

1/33.0%

3

5%

7

 

 

 

 

1/12.5%

1/4.7%

 

 

2

4%

8

 

 

1/33.3%

 

 

 

 

 

1

2%

9

 

 

1/33.4%

 

 

 

 

 

1

2%

10

 

 

 

 

1/12.5%

 

 

 

1

2%

 

Source: Frost & Sullivan and Airclaims

 

In spite of the problems created by the introduction of new technologies in the latest generation of satellites, there is no going back. It is commonly felt that the major issues in this generation of satellites have been dealt with and that the fixes have been largely successful. Effectively, the satellite failures of the late 1990s were a massive “beta test” in which incompletely developed technologies were tested by actual users. This is a common method of testing software, but was not an appropriate way (however unintentional) to test satellites in the environment of space.

Figure 9: Annual Anomalies by Value of Claims

Year

Antenna

Battery

Control Processor

Payload Electronics

Propulsion

Solar Array

Structures

Transponder

Claims (US$M)

% of Total Claims

1

279.4/100%

31.2/6.1%

 

138.0/90.1%

67.0/13.8%

955.3/45.0%

12.4/100%

19.0/52.8%

1502.3

36%

2

 

311.2/60.8%

 

 

174.0/35.8%

567.1/26.7%

 

 

1052.3

25%

3

 

169.3/33.1%

 

15.1/9.9%

 

347.3/16.4%

 

 

531.7

13%

4

 

 

 

 

123.4/25.4%

215.0/10.1%

 

 

338.4

8%

5

 

 

 

 

 

8.9/0.4%

 

 

8.9

<1%

6

 

 

180.0/30.9%

 

20.0/4.1%

 

 

17.0/47.2%

217.0

5%

7

 

 

 

 

101.0/20.8%

30.0/1.4%

 

 

131.0

3%

8

 

 

270.0/46.4%

 

 

 

 

 

270.0

7%

9

 

 

132.4/22.7%

 

 

 

 

 

132.4

3%

10

 

 

 

 

0.1/<1%

 

 

 

0.1

<1%

 

Source: Frost & Sullivan and Airclaims

 

Regardless, the testing has occurred, problems have been identified, and fixes have been made. After this massive investment (paid for by underwriters, operators, and manufacturers), the number of on-orbit anomalies should decline, although they are doing so very slowly (and will never reach zero). However, it is not unreasonable to consider these failures in future procurement decisions. There is no guarantee that all of the systemic problems have been dealt with or that new problems will not arise. The record of failures in the late 1990s speaks as much to organizational as to technological issues.

return to top

The Next Generation of Large Satellites

After the failures of the late 1990s, there is some cause for concern about the upcoming generations of larger satellites. The EADS Astrium Spacebus 4000 and Alcatel Eurostar 3000 have been deployed (or are about to be). The first Eurostar 3000 is in orbit with no reported problems and the Spacebus 4000 should see service within the next year or so. These buses are roughly similar to the BS 702 in capacity and in growth over their predecessors. As with any new satellite, it will require more satellites and more time to fully prove the Eurostar 3000. The Spacebus 4000 will require the same sort of demonstration. Following the trials of the past few years, an operator buying one of these buses must feel a certain amount of concern.
Both the Eurostar 3000 and the Spacebus 4000 are developments of previous systems. The SS/L 20.20 and EADS Astrium/Alcatel Alphabus are entirely new designs. As such, there is the ever-present danger of failures caused by this lack of heritage. With this in mind, it can be expected that the first customers for these buses will exercise great watchfulness and pay high insurance rates (if they can get insurance at all). The operators who first order these satellites will be those that truly need their extended capacities and not those who simply want a few more transponders. The disastrous introduction of the BS 702 serves as and example that there is considerable risk and no guarantees in being among the first to use a new satellite.

return to top

Impact of Systemic Anomalies on the Space Insurance Industry

Since the late 1990s, the space insurance industry has been battered by a series of satellite and launch vehicle failures. In part, the magnitude of these losses was the result of the success of the insurance industry previously. The profitability of the space insurance industry in the early 1990s caused new players to enter the space insurance business. As the number of insurers increased faster than demand for insurance, the result was more money chasing a limited number of underwriting opportunities. As a result, rates fell and terms eased.
By the late 1990s, it became possible to insure a satellite’s launch and up to five years of on orbit service at one time. In-orbit insurance rates decreased to between 1 and 1.5 percent per year and the size of insurance package that insurers could take on increased dramatically (in 1998 the maximum available capacity for a launch was just under 1 billion dollars). Launch insurance for satellites on proven vehicles cost less than 15 percent, including five years of on-orbit insurance.

Figure 10: Underwriting Losses/Insurance Industry Profitability

Source: Frost & Sullivan

These insurance market developments seemed reasonable at the time they occurred. However, when satellites (and entire families of satellites) begin to develop systemic problems, broadly written insurance policies and optimistically low underwriting rates resulted in sizable losses for the insurance industry. The most obvious response has been an increase in premiums, with a current average rate for in-orbit coverage for a reliable satellite around 2.5 percent per year. There have also been a number of changes in the ways in which policies are written. These changes can be divided into two basic categories. The first are changes that reduce the time lag between anomaly events and insurers’ response to them. For initial launch and one year of on-orbit coverage, insurance policies are now written no more than six to 12 months before launch. This reduces the danger that an insurer will be unable to respond to adverse events, by limiting the period in which an anomaly can occur after a policy has been signed. Following this same logic, renewals of this initial on-orbit policy are limited to a length of one year and are issued within a month of their effective date, with each renewal treated as a new policy, with corresponding due diligence by the underwriter.
Launch insurance now costs 18 to 20 percent for a launch and one year of (more limited) in-orbit insurance for a reliable satellite and launch vehicle, and is calculated on the basis of at least half of the risk coming from the satellite. (In 1998 the split was 25 percent satellite, 75 percent launch vehicle).
Underwriters have also begun to tailor coverage to reduce specific known risks. Insurers now exclude coverage on satellite systems with known problems (either individual failures or systemic problems) and are reluctant to cover systems subject to on-orbit single point failures that would severely impact satellite function. If it were possible to get insurance on such problematic areas, the cost would be considerably higher than average rates (even at today’s higher levels).

Figure 11: Satellite In-Orbit Insurance Rates and Coverage, 1999-2004

 

Average Rate*

Very Good Rate**

Period of Coverage

Use of Exclusions

Early 1999

1.0% to 1.5%

<1%

up to 3 years

Not Used

2004

2.4% to 2.6%

1.6%

1 year

Common

* Satellite in good condition
** Satellite bus with an excellent record

Source: Frost & Sullivan

The other tailored coverage strategy (used by both insurers and operators) is to insure assets in such a way as to reduce the potential payout. This can be done through the use of large deductibles or by writing policies for less than a satellite’s full value (for instance 75%). Both strategies are increasingly being applied, as insurers and operator seek ways to provide at least partial protection against risk at an affordable price. While a few large operators are self-insured, most operators still minimize risk by investing in some amount of insurance.
In addition to increased on-orbit rates and more limited coverage, underwriters have also changed the usual definition of a “constructive total loss” from 50 to 75 percent of the spacecraft capacity, reducing payments for satellites that retain considerable functionality.
With these policy changes and a decrease in satellite failures over the past couple of years, the insurance industry has stabilized and is approaching profitability. This is important because, as has been noted, most operators are not willing (or able) to self-insure and thus the space insurance industry is necessary for the continued health of the satellite industry. At the same time, however, this stabilization has come at some cost. The increased expense of using insurance as a risk mitigation tool and the reduction in the amount of risk it protects against have made the avoidance of satellite failures even more important than it had been.

return to top

Conclusions

Over the past five to six years, satellite reliability has become a number one priority for operators, manufacturers, insurance brokers and underwriters. The unfortunate experience of this period has altered industry priorities. Operators such as PanAmSat will not quickly forget the trauma of spectacular losses such as that of Galaxy 4. As insurance costs have increased and become a much more significant component of the total cost of ownership of a satellite, operators have begun to factor insurance costs into purchase decisions and have increased their oversight during the manufacturing process, and particularly on the introduction of any new technologies. Where a decade ago, operators accepted new technologies willingly, in a quest for increased capability, there is now a preference for products whose reliability record is strong, and that have not suffered systemic anomalies. Even when purchasing from manufacturers with good reliability records, operators have insisted on the use of proven components and subsystems with a strong in-orbit flight heritage.
Likewise, the underwriting community that wrote policies for Boeing 702 satellites - and then received claims on 75 percent of the on-orbit fleet - have become considerably more cautious in their underwriting policies. The huge losses suffered by the underwriters in space insurance have resulted in the exit of a number of important underwriters from the space insurance sector and considerably reduced industry underwriting capacity while increasing insurance costs to satellite operators.
With every failure, satellite manufacturers have become more concerned about quality issues they might have considered well in hand. Satellite manufacturers interviewed for this study said that they were now more willing to delay deliveries to guarantee quality. Considerable effort has also been devoted to redesigning problematic satellite systems to prevent such massive failures in the future. As a direct result of the increase in number of satellite anomalies and a concurrent reduction in available underwriting capacity and breadth of coverage, spacecraft reliability has replaced increased capacity as a manufacturer’s most important benchmark.
It must be remembered, however, that there will always be anomalies and no satellite bus is perfect. The 2003 antenna issues with e-Bird 1 (on the most reliable bus, a BS 376) show that it is not totally free of problems. The satellite manufacturing industry has been forced to refocus on quality, having, in some cases, pushed its methods and technologies farther than they would go, but this is a natural (if undesirable) part of the industrial product cycle. The future will demonstrate who has best understood the lessons of the last few years, by retaining or regaining a proper balance of heritage and innovation and by cultivating a culture of quality so that they will become the standard by which others are judged.

return to top

Frost & Sullivan
2400 Geng Road
Suite 201
Palo Alto, California, 94303
USA
tel: +1 650 475 4537
fax: +1 650 475 1571
http://www.frost.com/

 

Naam auteur: B.T.C. Zandbergen
© 2014 TU Delft

Metamenu