Space Shuttle

The Space Shuttle Columbia Disaster: Clarity of Purpose and its Impact on Risk
by Frank Winters, first published on the Gantthead website

Introduction

Space Shuttle Mission STS-107 ended in disaster on February 1, 2003, cutting short the lives of the seven remarkable crew members. The United States and most of the world is in mourning over this tragic event. Once again, the United States space program is being subjected to criticism, costly delays and close scrutiny.

One of the factors contributing to controversy over the Space Shuttle Program has been a perceived lack of clarity of purpose. This lack of clarity has made it difficult to determine an acceptable level of risk. We have heard many people speak out in favor of continuing the space program. Retired astronauts, politicians, NASA administrators and journalists have said we must continue space exploration because mankind will always seek to understand the unknown, will continue to push the envelope and will continue to reap many benefits from this important work. Most advocates say we know there are risks but we are willing to accept risk given the importance of the mission. But what is the risk, what is acceptable risk and what exactly is the mission?

Where does the execution of the Space Shuttle Program stand? To quote NASA’s website, “STS stands for ‘Space Transportation System’, a.k.a. the Space Shuttle, and ‘107’ is the ‘flight tail number’ or the 107th flight of the Space Shuttle, although the order of missions may have changed after assignment of the flight number.” So by nomenclature design, STS-107 should have been the 107th flight, but in practice that’s not exactly correct. According to an article in Sunday’s online addition of the New York Times, there have actually been a total of 113 flights of the Space Shuttle. (Other articles say there have been 111 flights.)

Therefore, the tragic ending to STS-107 is the second catastrophic result out of about 113 flights.

Many questions arise: Was this latest disaster an unexpected accident, a function of the inherent risk of space travel, a result of human failure or an incident in the development of an experimental mode of transportation? Should we be shocked and surprised when a shuttle mission ends in disaster? Did the astronauts realize that the probability of catastrophic failure was so high? Is a 2-in-113 probability of loss of crew and vehicle an acceptable risk level, given the nature of the mission? Is this probability about what NASA expected?

NASA’s Mission Unclear

Acceptable risk level is a function, in part, of the purpose and importance of a program. There is reason to believe that the purpose and mission of the shuttle and space program itself is unclear. Sean O’Keefe is the administrator of NASA, having been appointed in December of 2001. The NASA website quotes O’Keefe: “In broad terms, our mandate is to pioneer the future...to push the envelope...to do what has never been done before.”

The National Aeronautics and Space Act, passed by Congress in 1958, states: “The Congress declares that the general welfare of the United States requires that NASA (as established by Title 5 of this act) seek and encourage to the maximum extent possible, the fullest commercial use of space.”

NASA’s earlier statement of purpose is quite clear. Administrator O’Keefe’s statement, while visionary, is much less clear. The shuttle program is an important part of the space program. Is the purpose and role of the shuttle program clear?

The Changing Purpose of the Shuttle

The shuttle was designed to provide a reusable means of transportation to get people into orbit. One stated goal of the shuttle program as presented to President Nixon in the early 1970s was to make space travel routine, with weekly trips and rapid turnaround. Prior to the Columbia disaster, some people in the program were saying that this goal had been met. Seventeen years later, with the launch of Columbia, it is apparent that the shuttle has not nearly achieved the reliability required of a routine mode of transportation.

Initially, the shuttle was to travel to a space station. However, the station was deleted from the original plans due to cost. When the space station was deleted, NASA sold Congress the idea that the shuttle could become a commercial enterprise, hoping to win the funding required to continue the shuttle program. It would be a primary means of doing work by contract such as launching satellites. It would in effect pay for itself. This never came close to becoming a reality.

NASA told Congress that the shuttle would fly weekly, at a cost of $5 to $10 million per flight. Today, the cost--as reported in this week’s Economist magazine--is $500 million per flight. Of course, the shuttle makes only five or six flights per year, and fees for commercial use of the shuttle never came close to the per-flight cost.

A Little History

The U.S. space program, along with those of other nations, has its origin in the early 1950s. In 1952, the International Council of Scientific Unions decided to establish July 1, 1957, through December 31, 1958, as International Geophysical Year (IGY) because cycles of solar activity would be at a high point during that time, making it an excellent opportunity for exploration and experimentation. As we know, Russia was the first country to successfully launch a satellite, on October 4, 1957, and they repeated the feat with Sputnik II on November 3 of the same year. The U.S. Congress passed the Space Act the following year, creating NASA and the modern U.S. space program.

Kennedy’s Mission Statement

Our space program and NASA got an enormous shot in the arm when President Kennedy set a goal to send people safely to the moon and back within the decade of the 1960s. This was accomplished more than once, with one disaster (the Apollo mission disaster that caused the deaths of three astronauts--Grissom, White and Chaffe--while still on the ground in 1967) and one near disaster (Apollo 13).

During this time and up until 1972, when the last moon shot took place, the U.S. space program had a clear mission and purpose. A book about the Mercury and Apollo missions is entitled Failure Is Not an Option. This title sums up the focus and attitude toward risk mitigation that results from a clear mission statement and sense of purpose.

Moon Shots End, Space Stations and Shuttle Gain Center Stage

The successful moon shots of the Apollo program ended with Apollo 17 in 1972. Since that time the direction and purpose of the space program generally, and the space shuttle program specifically, have been less clear.

In 1972 the funding for the space shuttle was approved--but not for the space station it was designed to serve. The shuttle was born with an unclear mission. In 1981, the first shuttle was launched. While it was conceived as a shuttle, it is very often described as an orbiter, partly because until the late 1990s it had no destination other than Earth's orbit.

In 1984, the concept of the International Space Station was envisaged, and in 1995 Russia, 11 European states, Japan, Canada and the United States endorsed a plan to build the station. In 1998, 15 nations agreed to work together to build the station, with construction starting that year.

The Shuttle and International Space Station (ISS)

The shuttle and International Space Station are now components of an approach to space exploration that has been discussed, designed and redesigned for many years. Today, however, the mission of the ISS itself is also more vague than it once was. Originally thought of as a way station for launches into deeper space (including Mars exploration), it is now the site of various outer space experiments.

In his Time magazine Viewpoint article entitled “The Space Shuttle Must be Stopped," Gregg Easterbrook says that the space station was created to provide a destination for the shuttle and the shuttle was created to provide transportation to the station.

What is the Mission of the Shuttle Today?

Space station technology has had a long history. The current design is primarily the work of western and Russian scientists. Collaboration between nations has continued. For example, between 1995 and 1998, U.S. space shuttles docked with the Russian MIR space station nine times. In fact, the day after the Columbia disaster, a Russian Soyuz shuttle took off to travel to the ISS to restock the station.

One can only wonder why there was a U.S. shuttle mission that did not (and, as it turns out, could not) visit the space station while at the same time a Russian craft was ready, willing and able to perform what was originally a primary function of the U.S. shuttle program. Today we must ask for clarification of the shuttle’s purpose. In fact, we must ask for a review of the overall plan for space exploration.

There are many questions in our minds about the shuttle’s purpose:

Is the shuttle a means of transportation designed to routinely carry civilians or military personnel on critical mission? Who will ride the shuttle?
Is it an orbiting laboratory or a transportation system?
What kind of experiments should be funded? To what purpose? Why can’t the experiments be done in the space station?
How frequently is the shuttle needed? When first conceived, it was thought that shuttles would fly to the space station once a week.
What size payloads does it need to deliver?

Risk and Probability of Failure

Lack of clarity of mission and purpose has made risk analysis very difficult. Time magazine this week says: “In a flying machine with more than 2.5 million parts, even a 99.99 percent reliability level would still leave 2,500 things to go wrong.”

I’ve seen other references to "four 9s reliability" attributed to the shuttle. These references are, of course, out of context and can’t possibility refer to the reliability of the shuttle as a system. Some people refer to four 9s and say it’s the best you can expect. This attitude contributes to an acceptance of an indeterminate level of risk, which can lead to the acceptance of inadequate systems and/or a lack of emphasis on risk mitigation.

What is the risk of catastrophic failure of the shuttle? According to an article published in the Boston Globe, referencing references the Congressional Record, the expected failure rate of the existing shuttle design is 1 in every 250 voyages. A recent Newsweek article says: “There was a time 17 years ago, before the space shuttle Challenger blew up, when NASA claimed the likelihood of a catastrophic accident was 1 in 100,000. Since that time, NASA has increased the risk to 1 in 148.” However, in a Sunday New York Times article, a different number is given: 1 in 78 before the Challenger disaster; 1 in 483 according to recent NASA statements.

A New York Times article quotes Theodore Postal, an MIT professor and expert on complex spaces systems, saying: “The most reliable systems or booster, outside the human space program, have about one failure in 50 tries, for a 98 percent success rate.”

According to the Globe article, a second-generation spacecraft has been under development. The second-generation system has a reported design objective of a failure rate of 1 in 10,000 missions. According to the Globe, $1.3 billion has been spent on the second generation program to date, but it is stalled at the present time with decisions regarding how to proceed delayed until at least 2005. It’s pretty clear that the failure rate NASA has expected over the years has itself been unclear and has changed along with the mission.

Another Globe article on Sunday entitled “Officials had warned of NASA safety issues” quotes an expert opinion that I believe is shared by many. The expert is Howard McCurdy, a social scientist at American University and author of six books about NASA: “The shuttle has always been a high-risk venture,” McCurdy writes. “You worry that it finally caught up with us: the fundamental design. There are a lot of components on the shuttle for which a component loss can very quickly cause a vehicle loss.”

Professor Postol of MIT is quoted further in the Times article: “You have a system where the consequence of what appear to be relatively minor failures quickly propagate into catastrophe. It looks like the system is roughly as reliable as most other launch systems of roughly comparable complexity.”

Professor Postol added: “You’re riding a stick of dynamite into space. We know how to do that, but sticks of dynamite can explode.”

These expert opinions and characterizations, coupled with the failure rates associated with the shuttle, add up to a design not safe enough to reliably transport people. The shuttle, many believe, is still an experimental mode of transportation.

Is Risk “Normalized” at NASA?

In 1996 Diane Vaughan, a professor of sociology at Boston College, published an analysis of the Challenger disaster. Her book, entitled The Challenger Launch Decision, Risky Technology, Culture and Deviance at NASA, concludes that the NASA culture allowed and even facilitated the acceptance of a known flaw in the rocket boosters as an acceptable risk when many of the engineers involved in the design of the boosters recommended against the continued use of the design.

The night before the fateful launch of Challenger, engineers from the contractor Morton Thiokol, to quote the book, “argued against launching on the grounds that the O-rings were a threat to flight safety. NASA managers decided to proceed.” The book points out that the O-ring problem was well-known for years; the relationship between temperature and failure had been established with low temperatures tending to cause failure.

The design was thought by many experts to be seriously flawed. Vaughan's research, which is extensive and completed over nine years, finds that at NASA, risk tends to be normalized. That is, when a risk is identified, the culture at NASA tends to work the risk into normal operational standards. In other words, NASA does not always work to redesign risky components. Whether this is an acceptable practice or not depends, in part, on the specific nature of the mission.

It appears that the NASA culture may not have changed enough since 1986. Today we don’t know what caused the Columbia disaster. However, the heat shield made of ceramic tiles is a chief suspect. They have been a known problem since the first launch in 1981, also a Columbia launch. During the launch of STS 1, Columbia's and the shuttle’s first launch, reports say that 12 tiles fell off the shuttle. If tile failure or damage turns out to be the cause of this latest disaster, history will have been allowed to repeat itself. Whether the disaster was caused by tile failure or not, the risks associated with travel on the shuttle appear to be far higher than an acceptable risk, given the shuttle’s almost routine usage.

This concern is heightened by the apparent lack of a rigorous response to the insulation that fell off Columbia on lift-off. During the flight of Columbia, no attempt was made to ensure that no damage was done by the falling insulation. In the first few days after the disaster, NASA explained this by saying there was nothing that could be done anyway.

No space walk to examine or repair anything was possible, no docking with the space station, either. If the ship was doomed at that point, nothing could be done. This seems incredible given what was at stake, the nature of the mission, and the dangers and known problems associated with the shuttle.

Clear Goals and Objectives are Needed

The national agenda has moved away from the space program. Many citizens and government leaders feel it is wasteful and not critical to our country’s well being. Explorers usually have a physical destination: North America, the moon, Mars or beyond. It’s been 30 years since anyone has traveled more than 300 miles from Earth. In hindsight, if our goals were to orbit the Earth, perform experiments and launch and use telescopes to visually explore space, we would have designed a different system than the one we have.

While it makes us feel good to say this tragedy will not deter us from space exploration, we must stop now and reconsider every aspect of our space program. Our objectives must be clarified in the context of the broader national agenda. Once we know exactly what is at stake, determination of acceptable risk targets can be made (the current ones seem to be too unfavorable by one or two orders of magnitude). Knowing the mission, purpose and acceptable risk will enable our scientists and engineers to design a system that will eventually perform as required.

Lessons Learned

NASA’s track record, given that it has undertaken some of the most complex and difficult projects and longest programs in history, is superb. However, we can see in hindsight that difficult, risky tasks require crystal clear statements of purpose and near perfect alignment within the team. At NASA, the culture of pride and dedication creates teams that do great work day-in and day-out.

What has been lacking is continuous reinforcement of the specific mission statement so that it is clear to all. Most IT projects have missions that are mundane compared to those of NASA. On such workaday projects, it is very easy to lose focus and forget what the project is about. Project teams very often stop questioning and stop seeking clarification. Leadership forgets to remind the team what the mission is. This can lead to project failure.

There are some lessons we can take away from the space shuttle’s difficulties:

Lesson No.1: Never stop questioning; be certain that you know how your work fits into the overall mission and purpose of your project. Read the scope documents every day. If no scope documents exist, get them built or leave the team. If they are unclear, get them clarified.
Lesson No. 2: Without a clear understanding of the purpose and mission of our work, we can’t know what risk probability is acceptable. This is incredibly dangerous! Every program and project has inherent risk. The purpose of risk mitigation is to reduce the probability of the risk to an acceptable level. We must know what level of risk is acceptable in order to know how much resource should be spent in risk mitigation.
Lesson No. 3: In very risky work, it is not helpful to continuously say, “risk is unavoidable.” There is a huge difference between rolling the dice and mitigating risk. If there is risk that can’t be mitigated in some way, the project plan must be questioned. For IT projects, this might mean (for example) admitting that the timetable is not feasible and negotiating for more time rather than leading the team on a death march.
Lesson No. 4: Question the impact of changes in requirements on design. The Space Shuttle was designed 30 years ago for a purpose that has changed. Is the design suited to today’s purpose? The same question needs to be asked throughout the life of any program or project.
Lesson No. 5: Know the culture of your company. How are risk and its mitigation usually managed? For example, if a risk is discovered, does the organization tend to shoot the messenger? Is there a tendency to rationalize issues such as risks or does the organization meet them head-on? Understanding the culture you operate in will help you manage risks and other project impacts.

Go to Current Events articles index