The SpaceX Starship SN10 prototype coming in for a landing during a flight March 3 at Boca Chica, Texas. (credit: SpaceX)

Putting the SpaceX-FAA dispute in context

by Wayne Eleazer
Monday, March 8, 2021

On January 25, 1957, the first Thor IRBM launch occurred from Launch Complex 17 at Cape Canaveral Air Force Station. The objectives of that first-ever Thor launch operation were modest: to proceed down through the countdown, load the liquid oxygen, and start the engine. Anything useful that occurred after that was pure gravy. As it turned out, contamination in the liquid oxygen led to a valve failure and Thor 101 barely rose off the launch pad before the engine quit and the vehicle fell back down, creating a massive explosion and damaging the launch pad. Nonetheless, since the objectives of the operation were all met, it was a “success.”

SpaceX has conducted a number of other similarly “successful” operations in the development of its Starship vehicle, launched from the company’s complex in Boca Chica, Texas. Most have resulted in massive explosions, but on each one something new and important was learned, and thus constituted a successful test. Some things cannot be simulated, and often a launch is the first occasion that all of the required elements actually interact.

With any launch there are certain hazards that have to be mitigated. Falling debris, which may contain firebrands that can have an incendiary effect, is the most obvious. Some vehicles can produce toxic clouds when they explode or burn. And when a booster blows its top, the explosion can produce significant shock waves that may not be strong enough to actually harm anyone standing in the open at some distance away but can damage structures even further away and produce casualties as a result.

The national standard for the acceptable risk posed to areas external to the launch site is a probability of 30 in a million. This originally was roughly based on the threat presented to the general public due to aircraft, the idea being that no additional risk beyond that already accepted from aviation should be allowed. Calculating the probability of casualty requires assessment of the nature of the hazard, the probability of a certain type of failure occurring in each portion of the flight, and the vulnerability of the population that can be affected. Vulnerability of the population depends on not only the number of people potentially at risk but also the type of structures they inhabit and even the time of day, since people are more likely to be inside structures at night. And just what constitutes a casualty is a complex issue that we can’t address here.

The primary means to mitigate the launch hazards is to locate launch sites away from populated areas and to limit the allowed trajectories. At Cape Canaveral, Vandenberg Air Force Base, and Wallops Flight Facility, the launch pads are located well away from populated areas, trajectories largely are limited to over-water flight paths, and evacuations are conducted on the launch base as required to limit risk exposure.

But it is not quite that simple. Weather conditions can greatly impact the area threatened by the launch hazards and that has to be considered in real time for each launch.

One of the hazards greatly impacted by weather is “distant overpressure focusing.” Under certain conditions the shock wave from an explosion can propagate well beyond the area that normally would be affected. In fact, the Texas City disaster of 1947 illustrated this effect. When a ship in port loaded with nitrate fertilizer exploded, resulting in the worst industrial accident in US history, the damage in some of the areas further away from the site was worse than areas that were closer. So, distant overpressure focusing has to be assessed at the time of the operation to determine its impact on the probability of casualty.

At the launch bases at Cape Canaveral, Vandenberg, and Wallops, there are government agencies charged with ensuring safety. The probability of casualty for an operation is assessed both during the planning process and in real time prior to the operation; go/no go decisions are made accordingly. Launches will not be allowed to occur if the 30 in a million requirement is exceeded. At Boca Chica, a privately owned launch complex, that safety task is accomplished by SpaceX, with approval by the Federal Aviation Administration by means of the analyses required by the launch license issued to SpaceX by the FAA.

On December 8, 2020. SpaceX planned to launch SN8 Starship mission from Boca Chica. SpaceX conducted the required flight safety analyses and found that the distant overpressure focusing probability of casualty limits would be exceeded. The company asked the FAA for a waiver of the requirement; the FAA refused. SpaceX launched SN8 anyway, and the vehicle was destroyed during the landing attempt.

We do not know exactly when this situation became known to the FAA, but about fove hours before the planned SN9 launch on January 28, the agency informed SpaceX that the launch was not approved. The FAA required SpaceX to conduct an investigation into the prior noncompliance and made changes to the launch license as a result. The SN9 mission was launched on February 2.

So, the FAA’s delay of the approval to launch SN9 had nothing to do with the fact that SN8 had crashed and exploded. SpaceX is free to blow up any number of its vehicles and create damage to any of its own facilities, but not free to exceed the national standards for the probability of casualties to the general public. Perhaps, if SpaceX killed off its entire launch site workforce in a mishap, then OSHA might get interested, but it would not be an FAA concern.

Elon Musk has complained that the FAA’s regulatory structure is “fundamentally broken” and called for revisions to the standards that he said had been established back when there were only a relative few launches each year from government launch ranges. But such fundamental revisions would require increasing the allowable casualties in the civilian population. Nothing else would have helped SpaceX launch in December, unless the distant overpressure focusing requirement itself was done away with.

While the current number of launches from US ranges may not be very numerous, in the late 1950s and 1960s multiple launches of missiles and space boosters from the ranges each day were common. But in that time frame the ranges actually tended to use “risk avoidance” rather than actual risk management and the basic approach often was to avoid creating the hazard at all. The hazards associated with distant overpressure focusing and of toxic plumes from burning solid rocket motor propellants were, at best, not well understood at that time. But if the hazards had been realized the safety approach would have been to simply not allow launches if those conditions existed.

Analytical methods have improved greatly since then and we have a much better understanding of the hazards involved and how to employ scientifically based risk management techniques and utilize computer models to better assess them in real time. It is not possible for us to step back to a time period when those kinds of hazards were not recognized simply to reduce possible launch delays. In fact, the risk management approach now utilized no doubt offers an opportunity for reduced safety-related restrictions on launches than the older, more conservative risk-avoidance approach.

Some have described the FAA as standing in the path of progress, but an analogy to the SpaceX situation would be the FAA deciding that the airlines could crash a certain number of their airplanes without the agency being concerned. No one ever suggests that, although some might even argue that is what occurred with the Boeing 737 MAX debacle. And the fact is, commercial space transportation is already just about the least regulated aspect of the FAA’s responsibilities (see “Space myths”, The Space Review, October 1, 2007).

Increased numbers of launches as well as launches from new sites, together with the much greater number of people residing near the existing launch bases, present challenges to the space launch industry At Cape Canaveral, for example, the Air Force has been willing to adopt autonomous flight termination systems that reduce the time required between launches as well as revised analyses that enabled the first polar orbit launch from the Cape in decades. Further improvements in launch availability may be possible, but ignoring the actual risks involved is not a realistic option.

Wayne Eleazer spent 25 years in the US Air Force, serving as the Thor program manager, GPS integration manager, Atlas test director at Vandenberg Air Force Base, and led the space launch section of the Air Force Acquisition Directorate of the Secretariat in the Pentagon. Prior to his retirement in 1999 at the rank of lieutenant colonel, he served as chief of advanced planning of the 45th Space Wing at Patrick Air Force Base, Florida, responsible for initial planning of EELV and other launch operations at Cape Canaveral.

Note: we are using a new commenting system, which may require you to create a new account.

Home