8
A Tough Place to Work
Any sufficiently advanced technology is indistinguishable from magic.
Arthur C. Clarke’s Third Law, Profiles of the Future: An Inquiry into the Limits of the Possible
At least 600 satellites were whirling in orbit above the Earth at the start of the year 2001, an armada of commercial, military, and civilian science spacecraft totaling about $100 billion in public and private investment. Nearly 100 of those spacecraft relay military secrets and strategies, everything from images of enemy armies and missile silos to weather reports from remote battlegrounds of the world. Another subset of those 600 satellites gathers scientific data for government agencies, watching cloud patterns, whale migration, ozone holes, and the exploding stars in far-off galaxies.
The rest belong to a burgeoning industry—satellite communications. Private firms having been using satellites for the delivery
of mass media for many years, with radio and television networks—even those “cable” networks—beaming their programs via satellite to the local dishes of their affiliates. Telephone and paging services are also traditional customers of the satellite industry. But today, satellites also relay stock indexes, transactions from automated teller machines, retail inventory counts, truck movements, and consumer payment information at some pay-at-the-pump gasoline stations. Satellites have become a major yet invisible part of the commerce of modern economies. Most of us don’t realize how much our lives are intertwined with satellites until our pagers go silent or our favorite television programs are blacked out. And most of us know even less about the intricacies of using satellites than we do about how oil, gas, wind, or coal gets turned into the electricity coming into our house.
It took more than three decades for humans to put those 600 active satellites into orbit, not to mention the thousands of spacecraft that have already expired. According to the Teal Group, an aerospace consulting firm, more than 2,100 new spacecraft have been proposed for launch by the year 2010, at a cost of at least $220 billion. About 1,000 of those payloads will be commercial communications satellites.
Yet just as all of these satellites are queuing for takeoff, the failure rate for launch vehicles and satellites has increased dramatically, according to Christopher Kunstadter, senior vice president of U.S. Aviation Underwriters, a major satellite insurance company. The long-term failure rate for all launches is roughly 10 percent, he notes, but in the first 10 launches of a new type of rocket, the rate jumps to 20 percent. And the problems don’t stop at the launch pad and rocket booster. Space weather has been blamed as the cause or contributor to more than $500 million in insurance claims between 1997 and 2001, and in numerous other cases, satellite companies lost some of their capacity and redundancy in mishaps that could not be claimed. Since January 1998, the insurers of satellites have paid more than $3.8 billion for 22 total mission losses and 91 partial losses; uninsured losses over the same period exceeded that amount. In the grimmest of recent years, 1998 and
2000, the value of insurance claims exceeded the amount of money collected in premiums.
“The space insurance market is reeling from recent losses,” notes Kunstadter. “These losses reflect a real decrease in space system reliability. As a result, available capacity [for insurance] has decreased, rates have increased for all types of coverages and risks, and general terms and conditions have tightened.”
Furthermore, commercial pressures in this fiercely competitive market are forcing companies to fly satellites that may not be as hardy as the satellites of the past. “Cost and schedule drive technical considerations,” Kunstadter says, so things like spacecraft testing are being reduced in order to meet schedules. The high cost of radiation shielding for satellite parts also means that—due mainly to financial pressures—most satellites are less sheltered today from space weather than they were in the past.
Most satellite failures occur early, either on the launch pad or shortly after leaving it; others fail due to the breakdown of key parts once the spacecraft is out of reach in space. Orbital debris and meteors take out a few satellites. But experts cannot agree on just how many perish due to space weather. The Aerospace Corporation, a leading space technology research firm in the United States, claims that about 5 percent of all spacecraft failures can be attributed to the environment; the cause of another 10 percent of mishaps is “unknown.” According to Alan Tribble, who has been teaching courses on space weather effects since 1992, that figure for environmental failures might be as high as 25 percent, if you account for the long-term degradation of solar cells and other spacecraft parts. Dave Speich, a recently retired forecaster for the National Oceanic and Atmospheric Administration’s (NOAA) Space Environment Center (SEC) estimates that one-third of all spacecraft anomalies—including failures and corrected problems— can be attributed to the brutal environment of space.
The trouble with assessing the impact of space weather is that without collecting the satellite and looking under the hood to see what went wrong—something that is logistically and financially impossible for all but a few NASA satellites—it is difficult to know
with certainty just what caused a given spacecraft to fail. Satellites deliver reams of data on the health of their electronic systems and the local space environment, and that information is scrutinized down to the millisecond. Satellite companies regularly reconstruct in fine detail the exact sequence of events before a spacecraft anomaly or failure. But such data and the analysis of it are closely guarded secrets in the satellite industry, proprietary matters that are rarely shared with scientists, much less competitors.
So whenever a satellite dies in the middle of a space weather event, many space scientists—and a fair number of engineers from competing companies—speculate about whether the environment had something to do with it. Maybe a storm wiped out a satellite or perhaps it was the trigger event that exposed a more fundamental problem with the design of the spacecraft (such as the “tin whiskers” of Galaxy IV). The predicament is that these outsiders—some of them with an agenda, others with a lifelong scientific interest—can only cite circumstantial evidence, like trying a murder case without a witness. Unless they work for the company that suffered the loss, they don’t have access to the data to prove how space weather may have killed any satellite. On the other hand, those same speculators are rarely refuted or challenged in public. It is a puzzling and intriguing dance between scientists who understand the space environment and engineers who know their satellites, between researchers bucking for the next publicly funded grant and researchers protecting their company’s future.
On January 13, 1994, the Sun began spraying high-speed solar wind at Earth. A hole in the corona was allowing the fastest breed of solar wind—usually two to three times faster than the everyday variety—to escape the Sun’s intense magnetic fields and rush across the solar system. At Earth a minor magnetic storm began and proceeded rather humbly until January 19. NOAA’s Space Environment Center noted the storm but did not feel compelled to issue any alerts. It was, by most measures, a relatively modest event.
Modest or not, it was enough to disable a satellite. At about 12:40 p.m. on January 20, the Anik E1 satellite began to spin out of control. Operated by Telesat Canada, the three-year-old, $300 million satellite lost contact with ground stations when a failure occurred in the primary attitude control circuit that kept the spacecraft oriented and pointed in its orbit over the Americas. Fortunately, controllers were able to resolve the problem within about eight hours by switching to the backup system. E1 was pointed back at Earth and ready for business.
Its twin satellite was not so fortunate. Within an hour after communication with Anik E1 was restored, Anik E2 went into a similar tumble. The attitude control system had a failure much like the one that had temporarily felled E1. But when controllers switched to the backup system for E2, they found that it was dead, too. Anik E2 was useless and engineers struggled for months to restore the satellite.
The failure of two of Canada’s prime satellites caused communications havoc across the country. Several hundred of Telesat’s corporate customers were affected by the outage, including stock quote services, retail sales inventory systems, and scientific data relays. Long-distance telephone service was cut off in the northern reaches of Quebec and Ontario and all of the Yukon and Northwest territories. But the hardest-hit sector was the media. More than 100 newspapers, radio stations, and television networks were sent scrambling for alternative links on January 20. Many newspapers were using satellites to distribute stories and photos to their affiliates and to relay their pages to printing presses around the country; suddenly they were using faxes and film. The Canadian Broadcast Company could not distribute its Newsworld programming to 450 affiliates, and most of its specialty cable programming went black for the day.
Within 24 hours of the failures, most of Telesat’s customers had some form of satellite service restored. Many of the customers of the E2 satellite were switched to E1, while others were temporarily moved to Hughes’ Galaxy VI satellite. Eventually Telesat leased space on AT&T’s Telstar 301 to satisfy all of the remaining
E2 customers. Seven months later Anik E2 was resurrected to useful life when engineers developed a complicated system for controlling the satellite’s thrusters from the ground. After all the damage was done, Telesat forfeited between $50 million and $70 million in lost revenues and recovery expenses. The Anik E2 satellite lost a year and a half of useful life: six months spinning out of control and one year taken out of its life expectancy due to the amount of fuel that now has to be used to keep the satellite pointed in the right direction. The operating cost for the rest of E2’s life grew by $30 million.
Spacecraft operators and the forecasters at SEC were initially puzzled by the Anik failures. Telesat claimed that the environment was “100 times worse” than anything it had seen before, but a first glance at the usual scientific data did not support such a claim. As magnetic storms and solar wind streams go, the conditions were more active than normal, but nothing at all like the major storm conditions of 1989. The intensity of the energy coursing through Earth’s magnetosphere did not seem to implicate space weather as the culprit.
“Within a day of the Anik failures, I was called by representatives of three different groups,” says Joe Allen, former head of the solar-terrestrial physics division of NOAA’s National Geophysical Data Center. “One was someone from Telesat Canada; another was an official in the Canadian Ministry of Defense; the third was a vice president of the bank that had financed the building of the two satellites. All wanted to know whether the cause of failure could have been ‘sabotage,’ coincidence, or a result of a disturbed space environment. The U.S. Air Force facility at Falcon Air Force Base in Colorado had already run an analysis of the event and concluded with ‘high confidence’ that there was a ‘low probability’ that the failures were not due to natural causes.”
“I brought down the data on my computer and looked at the energetic electrons for the previous week,” Allen recalls. “The daily variation of the high-energy electrons looked like a textbook illustration of a smooth sine wave, they were so regular. I called Dave Speich to talk about it and told him I didn’t believe such
satellite anomalies could happen without energetic electrons being involved. Dave looked at the data and then said ‘Oh, my goodness.’ We both noticed then that while the daily variation looked like what you would usually expect, it was two to three times higher than normal. This condition had persisted for almost 10 days. It was not a ‘storm’ per se, although a Canadian university group mistakenly announced that it was.” But there was little denying that the environment had somehow damaged the satellites.
In addition to the Anik troubles, the Intelsat-K international communications satellite wobbled a bit on the twentieth, and the GOES (Geostationary Operational Environmental Satellites) and several spacecraft endured minor problems that day. “There was just so much compelling evidence that the environment had to be involved,” says Dan Baker, a University of Colorado space physicist who has studied the Anik problems. Closer analysis in the months after the failures revealed that the amount of high-energy electrons drifting in the radiation belts (around the time of the failures) had increased by a thousand fold and stayed that way for nearly eight days. The high-speed streams of solar wind buffeted the magnetosphere to a point where it began accelerating the particles trapped in the Earth’s space environment. Though it was not necessarily the sort of change that could kill a satellite instantly, Baker notes, the sustained increase in “killer electrons” probably inflicted a creeping death. It is known as deep dielectric or “bulk” charging.
Dielectric materials are supposed to be a spacecraft’s best friend, and usually they are. But sometimes they foster the most shocking of space weather effects. Dielectric materials are insulators—that is, materials that prevent the flow of electricity. Dielectric materials are used to insulate circuits and cables from each other deep inside the machinery, making sure currents flow where they are supposed to and do not disturb other components. Even when some of the highest-energy electrons from space penetrate a spacecraft and
lodge themselves in the dielectric insulation, those electrons leak away over time at a steady rate that keeps the satellite safe.
But when a spacecraft is immersed in a long-lasting and intense bath of high-energy particles, it can be overwhelmed. Because dielectric materials are designed to keep out most electric charges, the particles that do manage to get embedded in the insulation are inevitably particles of the highest energy. The accumulation of excessive electric charge in dielectric insulation is extremely rare, but when it does occur, it is often catastrophic. The effect is comparable to water flowing into a bathtub. During normal flows, the drain keeps the water level low. But if the flow suddenly increases, the water level rises until a new equilibrium is reached between the flow of water (electrons) in and the flow out (leakage from dielectric materials or metal components to an electrical ground). If the flow gets very high and stays high for long enough, the tub overflows. If the electron flux is high enough and persists for long enough, then electrical discharges occur that cause satellites to misbehave and sometimes fail. Scientists who studied the space environment during the failures of the Aniks suspect that deep dielectric charging was the culprit. The environment around the satellites was so intensely charged for so long that the buildup of electrons in the insulation provoked debilitating sparks and shocks.
Whereas deep dielectric charging by electrons is the most insidious and perhaps most destructive space weather effect, the most common form of disturbance is the single-event upset (SEU). SEUs are typically caused by high-energy ions, which are 1,800 times more massive than electrons. Accelerated to damaging speeds in the radiation belts around Earth, these particles can pierce a spacecraft and its components, literally punching holes through them. SEUs usually involve changes to the memory of a satellite’s computer circuits, sometimes locking up the electronic brain temporarily in the same way a terrestrial computer can crash. Other times the high-energy particles cause satellite sensors to give false readings (see Figure 13). Statistics from NOAA-SEC show that more than 950 SEUs were reported during the last solar cycle (roughly 1985 to 1995), about one for every spacecraft in orbit per
FIGURE 13. An image from the Large-Angle Spectrometric Coronagraph (LASCO) on the SOHO satellite shows solar protons bombarding and overwhelming the spacecraft’s instruments. SOHO and other satellites occasionally lose the use of their cameras when swarms of solar particles “white out” the view like interference “snow” on a television. Courtesy of SOHO/ European Space Agency and NASA.
year (though a few satellites actually endured the majority of those upsets and many more SEUs likely went unreported).
The Hubble Space Telescope is a famous victim of single-event upsets, though many satellites (including the Global Positioning System for navigation) have experienced them. Hubble’s orbit regularly takes it through the “South Atlantic Anomaly” (SAA),
which is located over South America and the Atlantic Ocean. The SAA is a region where the Earth’s magnetic field takes a sudden “dip” in intensity. This allows high-energy particles from the Van Allen radiation belts to reach lower than normal altitudes, causing problems for any spacecraft that flies through the SAA. One instrument on Hubble cannot operate when the satellite is passing through the region because an SEU could reset the instrument to its highest voltage and perhaps shock the instrument to death. That detector must be shut off for a few minutes every time Hubble heads for the South Atlantic, about 7 out of 16 orbits per day.
In 1999, two brand new NASA satellites—the Far Ultraviolet Spectroscopic Explorer (FUSE) and the Terra earth-observing spacecraft—suffered from corrupted memory and single-event upsets due to high-energy radiation from the radiation belts, particularly the South Atlantic Anomaly. As with Hubble, detectors on the new spacecraft sometimes have to be shut down when passing through that high-energy region. “It’s a ferocious radiation environment up there,” said Johns Hopkins physicist and FUSE principal investigator Warren Moos in a newspaper interview. “But I think the choice of [computer] chip was not the best.” Tougher, “radiation-hardened” chips could shield the spacecraft’s memory for decades. But as with the rest of the satellite industry, NASA is on a tight budget and a tighter schedule than in years past. Deputy project scientist Kenneth Sembach told Space.com that the FUSE engineers “chose the best they could given the costs and the time constraints we were under.”
Another space weather problem is known as surface charging, which can lead to a damaging “electrostatic discharge” on the outside of the spacecraft. As a satellite drifts through the plasma trapped in the space around Earth, it tends to accumulate electrons on its exposed surfaces. It is much the same as the way you pick up extra charge and shock yourself when you drag your feet across a carpet in the wintertime. Many spacecraft are designed with conducting surfaces to bleed off that accumulation of charges or direct it into grounding devices. But sometimes the flow of electrons is too much for the safeguards. The satellite lights up with
an electrostatic shock—space lightning—that can burn out a power supply or shock the components connected to the surface of the spacecraft. Sometimes these sparks can look like an electromagnetic signal from the ground and trick the satellite into following a phantom command.
In 1991 a potent electrical shock to the solar panels crippled the 10-year-old Marecs-A navigation satellite as it hovered over the Atlantic. In the case of another spacecraft, one of NOAA’s Polar Orbiting Environmental Satellites, the official cause of death was listed as a bolt sticking out of the solar panels. More than just a symbol of poor design and construction, the bolt became a sort of attractor for electrons. The bolts holding the solar panels together were supposed to have been insulated with a layer of Teflon. But the bolts were too long, so they pierced the Teflon coating during construction. After launch, the satellite flew without incident for two weeks until the radiation belts grew active. As electrons accumulated on the bolts, a lightning-like shock short-circuited NOAA-13’s solar power systems.
The energetic particles floating in the space around Earth also can damage spacecraft without causing immediate catastrophic failures. Instead, the destruction is slow, cumulative, and permanent. The steady dose of radioactive particles can destroy the principal energy source for most spacecraft: solar panels. Designed to absorb sunlight and turn it into cheap power for spacecraft, solar cells also collect a lot of other radiation from flares, coronal mass ejections (CMEs), galactic cosmic rays, or just the steady streaming of particles trapped around the Earth. High-speed particles smash into the semiconductor materials and reduce the expected lifetime over which the solar cells will be able to produce energy.
In October 1989, following a particularly intense flare that sent a swarm of high-speed protons toward Earth, at least 13 geosynchronous satellites suffered permanent damage to their solar cells, and one of the GOES weather satellites lost six years off of its designed life. A comparable storm in 1991 took two years of life from each of NOAA’s three GOES weather satellites by reducing their power-generating capacity. Over the course of its 15 years in
orbit, the Mir space station suffered power shortages due to the decay of its solar panels. And as recently as 1997, two newly designed satellites lost nearly one-fifth of their power-generating capacity after just two solar storms.
Space weather also has an indirect effect on satellites through increased atmospheric drag, or friction. Intense space weather events can heat the upper atmosphere of Earth, inflating it so that the gases around satellites in low-Earth orbit (LEO: altitude 300 to 500 kilometers, 180 to 300 miles) become denser. This expansion of the atmosphere significantly increases the number of microscopic collisions between the satellite and the gases and plasma of the upper atmosphere. The increased friction, known as “satellite drag,” can alter an orbit enough that the satellite is temporarily “lost” to communications links. It can also cause the premature decay of the orbit; that is, it can cause a satellite to fall out of orbit and burn up in the atmosphere sooner than planned. The Hubble Space Telescope has been boosted back up to a higher orbit during the servicing missions in order to counteract atmospheric drag. On the other hand, atmospheric drag from solar maximum 21 caused NASA’s Skylab space station to crash back down to Earth in 1979 before the first space shuttle could be launched to boost it into a more stable orbit.
The U.S. Department of Defense (DoD) spends roughly $500 million per year to mitigate the effects of the space environment. Those effects range from outright failures, to the loss of service time due to satellite degradation, to time spent correcting computer failures or satellite positioning. From 1987 to 1991—the height of the last solar maximum—DoD had to respond to as many as 300 satellite “anomalies” per year. Even in the years of solar minimum from 1994 to 1996, the military still dealt with 150 satellite problems per year.
With more than 30 communications and missile-warning satellites operating in geosynchronous orbit, 25 GPS (Global
Positioning System) spacecraft flying through the radiation belts, and 25 weather and science satellites flying in low-Earth orbit— not to mention the many classified reconnaissance spacecraft—the U.S. Department of Defense has strong reason to be concerned about space weather. In preparation for the current solar maximum and anticipating future space-based defense systems, the U.S. government’s National Security Space Architecture office gathered scientists, engineers, and decision makers from all branches of the military, civilian science agencies, and industry to conduct a national Space Weather Architecture Study. The group surveyed existing science and satellite data and projected that satellite operations could be disrupted as much as 15 percent of the time during solar maximum (2000-2002), with radio transmissions intermittently interrupted about 20 percent of the time. They also predicted GPS navigation errors about 20 percent of the time.
The vulnerability of the military is not necessarily an issue of hardware. Though some military satellites are using newer, less radiation-hardened parts, most of the crucial DoD satellites have been built to withstand the radiation of a nuclear weapon exploded in space. “Space weather is a threat, but not a day-to-day threat,” notes Chris Tschan, a retired Air Force officer who commanded the Air Force’s 55th Space Weather Squadron for a few years in the 1990s. “It is more of a nuisance to the military.”
The real problem is ignorance. Many of the officers and enlisted soldiers operating DoD spacecraft, communications devices, and navigation systems have scarcely heard of scintillation, CMEs, or solar wind. The Space Weather Architecture Study revealed that military users frequently don’t understand space weather effects and their operational impacts are not well documented for the average soldier. Out of those two problems spring a host of others, such as the fact that the current military requirements for space weather information and forecasts are outdated, fragmented, and incomplete. Military and civilian space weather requirements are similar but often addressed independently. And the lack of fundamental knowledge about the physics of space
weather and its impact on technology makes it difficult to develop and evaluate techniques to compensate for space weather.
“In the 1990s I observed that space system operators did not understand the magnitude or manifestations of the space environment on the systems they operated,” Tschan recalls from his days in the Air Force. “Most had a one-page summary, circa the 1960s, that told them there could be impacts. Since then I think the level of awareness of space environmental impacts is up substantially, but I feel confident that operators are still weak in knowledge about how to handle a space environment-caused anomaly.”
Tschan remembers an occasion when there was a partial radio blackout and false missile launch indications on the warning radar systems. The signals were caused by aurora, but the radar operators had no idea that the northern lights could cause such a signal. So they essentially ignored radar signal returns from that portion of the radar display, even though Air Force One was in flight at the time.
“An Air Force officer once remarked to me, ‘I can’t remember losing any spacecraft to space weather,’” Tschan recalls. That’s because many of the long-term effects of the space environment are hidden by the fact that DoD replaces many spacecraft before they have a chance to fail. “Basically, the space operators compensate for failing systems and switch to redundant systems, or they nurse ailing space systems along more often than you’d care to know, masking the problem to some extent.”
Lack of information about space weather is making the military vulnerable in other ways, too. As the federal budget shrinks and the military tries to get more for its dollars, the Pentagon is leasing time on commercial spacecraft and buying more satellites with off-the-shelf technology—that is, satellites that commercial firms already produce, instead of custom-made spacecraft and components. At the same time, only two companies are producing radiation-hardened parts, down from 20 suppliers in the Cold War era.1 Designers are installing less costly commercial computer processors in missiles and satellites because they weigh and cost much less than the bulky, radiation-hardened chips that DoD has
used in the past. Many officers who command or rely on space technology probably do not even realize that their budget savings may have a cost later. Others may be aware but do not have the money to do anything about it.
“I believe that military leaders are more aware of space environmental effects than they have ever been,” said Tschan. “But my feeling is that it isn’t enough, especially for systems with smaller-scale electronics that haven’t weathered a solar maximum before. Space weather is still something they don’t really understand because it isn’t tangible, like a thunderstorm or tornado. I’m sorry to say, it may take a small catastrophe for the vulnerability issue to sink in.”
On January 6, 1997, the Sun blasted a coronal mass ejection in the direction of Earth. By January 10 the cloud from the Sun was compressing Earth’s magnetosphere, stirring up energetic particles in the radiation belts and ionosphere and causing a major magnetic storm. By January 11 a telecommunications satellite died. Were the events connected?
AT&T announced to its customers and the news media on January 11 that its $200 million Telstar 401 satellite had “experienced an abrupt failure of its telemetry and communications” at 6 a.m. that morning. The satellite’s main control systems suffered a catastrophic failure, and the satellite went “out of orbital alignment,” the official statements noted. Four of the major television broadcast networks in the United States—ABC, Fox, UPN, and PBS—lost the ability to transmit programming to their affiliates. For Fox the failure couldn’t have come at a worse time, as the network was preparing to broadcast the NFC football championship the next day. Every PBS station in the country had received its programming via Telstar 401, so every station had to readjust its satellite dishes as AT&T found new relays for their signals. The satellite also was the primary relay for close to half of the syndicated television shows at the time, including “Oprah,”
“The Simpsons,” “Wheel of Fortune,” “Baywatch,” and several others.
Part of the U.S. earthquake-monitoring network also was disrupted by the loss of Telstar 401. The U.S. Geological Survey (USGS) had used the satellite to transmit near real-time data about seismic activity to the agency’s field centers. Though the regional earthquake networks in the more active regions—California, the Pacific Northwest, Alaska, and Hawaii—were not affected, almost everything east of Nevada was knocked out of service. USGS acknowledged that its reporting time for earthquake data for twothirds of the country was slowed by several hours.
The failure of the satellite ultimately turned out to be more of an inconvenience than a catastrophe for the media outlets. While operators worked to restore contact with the silent satellite, AT&T quickly transferred service for most customers to another satellite, Telstar 402R, as was spelled out in contractual contingency plans. Some customers, such as USGS, had to wait a bit longer, as their contracts were considered lower-priority items. Within six days, AT&T declared that Telstar 401 was “permanently out of service,” and AT&T Skynet began making plans to move another satellite, Telstar 302, into the defunct satellite’s slot in space.
The loss of Telstar 401 was a bit more than just an inconvenience for AT&T. In the fall of 1996, AT&T had agreed to sell its Skynet Satellite Services to Loral Space & Communications for a total of $712 million. With the loss of Telstar 401—which had been expected to last for another nine years—AT&T was forced to renegotiate. When Skynet finally changed corporate hands in March 1997, Loral paid $478 million. AT&T also collected $132 million in insurance claims for the loss of the satellite.
Lost in the shuffle of inconvenienced customers and humbled executives was the true fate of the satellite itself. AT&T announced at the time that a team of spacecraft experts from AT&T and Lockheed Martin, the manufacturer of the satellite, was being assembled “to determine the root cause of the problem.” But with Loral planning to build its own satellites and AT&T getting out of the business, neither company had a strong interest or obligation
to report on the results of the investigation. Louis Lanzerotti, a physicist for AT&T’s research cousin, Lucent Technologies, told Science News at the time that the satellite could have failed for several reasons, but “the coincidence with the magnetic storm is uncanny.” Since that time, nearly every scientist in the space weather community assumes and asserts that Telstar died from the one-two solar punch of January 1997. But none of them really know for sure.
There is an extremely low rate of failure due to space weather, at least by official counts. “Hundreds of satellites have functioned normally for close to a decade, providing an incredible financial rate of return for their operators,” notes one industry insider who prefers to remain anonymous. “The only reason that spacecraft failures make the news is that they are rare to the point of being shocking when they occur. The interruption in cable TV service due to power outages from terrestrial storms—or even something as mundane as a backhoe operator accidentally cutting a cable— never makes the news because we are familiar with and even expect bad weather and its inconveniences.”
But some scientists wonder if the space weather problems are rare or just rarely reported. Satellite operators and builders are under tremendous economic pressure not to report problems. Satellite builders certainly don’t want to scare away future customers or investors by admitting they might be vulnerable to space weather. Anomaly reports are privileged private information for the companies that build and own the satellites. Most of the parties involved—even the insurers—are bound by confidentiality agreements. “If the news gets out that a satellite might have failed due to space weather, that makes customers extremely nervous,” says one industry source. The cause of a failure is not necessarily information that companies care to share with investors and customers—and competitors—in a fiercely contested market.
“They don’t want to admit that there is a problem because it would create a competitive disadvantage,” says Joe Allen, who has long been the scientific community’s unofficial collector of anecdotes and data about spacecraft troubles. “Industry knows a great deal about the conditions in space and what happens. They say, ‘We know there are problems. We just don’t talk about it.’ It would affect the cost of their vehicles.”2
Some failures have been reported, principally the failures of government satellites. According to the Space Weather Architecture Study, 15 satellites have failed as a direct result of space weather since 1983, and those failures are spread out across the entire solar cycle, from active solar maxima to relatively quiet minima. Eight of those losses befell the first satellite in a new series, underscoring the unknowns in the development of new spacecraft. Chronic degradation by space weather also led to substantial redesigns of at least 12 major satellite systems in the past 20 years, and at least 21 satellites have had power supply problems that were caused by the environment. The authors of the architecture study summed it up for the skeptics in three words: “Silver bullets happen.”
Some industry watchers believe there could be a lot more silver bullets in the next few years. The race to claim a share of the space market has led many companies to build satellites with the newest untested technology or with less radiation hardening, which can double the cost of a satellite. “Budget and competitive pressures, as well as increases in demand for improved coverage, timeliness, accuracy, and assuredness in space-based service, are likely to make future space weather impacts more significant,” the authors of the architecture study concluded. Add in that the number of reported satellite anomalies traditionally doubles in years of major magnetic activity, according to Allen, and the current solar cycle could get interesting.
“A catastrophe is more likely for civilian systems since there are more of them and most have newer designs,” Chris Tschan notes. Most spacecraft manufacturers “overengineer” their systems because they don’t know exactly how much radiation their
satellite will face. In addition, spacecraft components tend to be tested individually for radiation hardness. But those parts are not necessarily tested in the context of an integrated, complete system. Then there is the problem of protecting a spacecraft from radiation. It costs about $40,000 for every pound that gets launched into orbit. Any radiation shielding beyond the minimum required to protect parts for the forecast lifetime is considered dead weight. Spacecraft designers are conscious of the need to squeeze the fat out of their designs, so radiation hardening seems to be a luxury most satellite builders cannot afford.
Building a hardy spacecraft is like building a bridge or a house to withstand an earthquake, says one industry engineer. “The daily space weather is not so critical. It’s the extremes that are important.” But as in the military, many of the people who design or operate commercial spacecraft today have never seen the extremes and scarcely know what to look for. The models of Earth’s radiation belts used to calculate environmental radiation doses are 20 to 30 years old. And the designers have so few case studies of failures or “anomalies” to study, unless their company has been one of the unfortunate victims of a storm from the Sun.
“I think space weather is a credible problem but not the only one to worry about,” notes Alan Tribble, author of one of the only textbooks on space weather effects. “I don’t feel that there are many glaring vulnerabilities in the average satellite, whether commercial or military. And as a rule, I think space weather does get appropriate attention in the industry. But most companies are prepared for the ‘usual’ environment. It’s the once-a-decade kind of space weather phenomena that always give problems, and it’s the same way on the ground. Hurricane Floyd [September 1999] caused problems on the East Coast because it was a very rare phenomenon. Things like that happen only a few days every few years. You either take your chances, or you drive up the cost dramatically by trying to make things hardened to survive the possibility of a worst-case environment.”
“It is generally more cost effective to offset satellite problems with redundancy than to try for absolute perfection,” notes one
industry insider. “Satellite operators will purposely carry extra capability on the fleet of satellites they operate, so if failures occur they can shift traffic to backup capacity on other satellites. We all have spare tires in our cars since no one knows how, at a reasonable cost, to build a flat-proof tire that lasts 40,000 to 80,000 miles.”
“It’s very difficult to quantify how susceptible things are to space weather failure,” Tribble adds. “I believe it boils down to a business decision. Solving the space weather problem has a cost associated with it. The return needs to exceed the investment. It’s a matter of risk.”
The simple fact is that no matter how expensive the hardware, no matter how commonplace the use of satellites, it will always be difficult and expensive to work in space. So is it negligent or just good business sense to take some risks with space weather? Satellite builders and operators are likely to find out through experience in the next decade.
