In September 1996, the Research Triangle Institute delivered a final technical report to the United States Air Force outlining methods for calculating the risks posed by unlikely but catastrophic space-booster failures. The report, prepared under Air Force contract FO4703-91-C-0112 and designated RTI Report No. RTI/5180/77-43F, was addressed to the safety offices of the 45th Space Wing at Patrick Air Force Base in Florida and the 30th Space Wing at Vandenberg Air Force Base in California — two of the country's primary launch corridors.
The document was authored by James A. Ward, Jr. and Robert M. Montgomery of RTI's Center for Aerospace Technology, Launch Systems Safety Department. Its distribution was restricted to U.S. government agencies and their contractors, with any outside requests to be routed through the 30th or 45th Space Wing safety offices directly.
What the Report Was Designed to Do
The core technical problem the report addresses is one of statistical edge cases: how do you build a defensible risk model when the failure events you're trying to account for are genuinely rare? Space-booster failures don't happen often. When they do, the consequences — debris fields, toxic propellant dispersal, potential casualties in populated zones downrange — can be severe. The challenge, from a safety engineering standpoint, is that standard probability models tend to perform poorly when applied to events with very low historical frequency but very high potential impact.
Patrick AFB and Vandenberg AFB both sit adjacent to civilian infrastructure. Patrick borders the Florida Space Coast; Vandenberg launches vehicles over the Pacific, with trajectory corridors that require continuous safety coordination. For the Air Force Space Command safety offices overseeing these ranges in the mid-1990s, having a rigorous analytical framework for low-probability, high-consequence failures wasn't academic — it was operational necessity.
The report was marked with distribution restrictions protecting what the Air Force classified as administrative and operational use data, a designation that limited its circulation while stopping short of full classification. The notation "DTC QUALITY INSPECTED" on the cover page reflects the Defense Technical Information Center's document processing, suggesting the report entered the broader defense research archive system upon completion.
Context: Launch Safety in the Post-Cold War Era
By 1996, the U.S. was navigating a complicated moment in its space launch portfolio. The commercial launch industry was beginning to expand, government payloads remained substantial, and the question of how to manage public safety risk at major launch facilities was becoming more pressing — not less. Both Patrick and Vandenberg were handling an increasing diversity of boosters, from legacy systems to early commercial vehicles.
Risk modeling of the kind described in this report sits at the intersection of actuarial science, aerospace engineering, and public safety policy. Getting it wrong has consequences that extend well beyond the range itself. Underestimating the probability of certain failure modes could mean inadequate safety corridors, insufficient public notification protocols, or flawed decisions about when to scrub a launch.
The fact that the Air Force commissioned an entire dedicated study on how to model unlikely failures — rather than simply applying standard statistical tools — suggests the existing frameworks were considered inadequate for the specific characteristics of space-booster risk. Rare events, by their nature, resist easy quantification. A booster that has flown a hundred times without incident provides limited statistical evidence about what happens on flight one hundred and one.
What the Document Tells Us — and What It Doesn't
The public version of this document, as preserved in the available record, provides the administrative framing of the report but does not include the full technical methodology or findings. The cover page, report documentation form, and distribution restrictions are legible; the underlying modeling frameworks, data inputs, and conclusions remain unavailable in this excerpt.
What is visible is enough to establish the report's legitimacy and institutional weight. A multi-year Air Force contract, a named research institution with an established defense research track record, named authors with specific departmental affiliations, and two major Space Command safety offices as the intended recipients — these are the structural markers of serious technical work, not a peripheral study.
Whether the specific modeling approaches developed here influenced subsequent range safety doctrine at Patrick or Vandenberg isn't something the available record resolves. What it does confirm is that in the mid-1990s, the Air Force was investing institutional resources in understanding the limits of its own risk calculations — a posture that carries implications for how launch safety decisions were being made during a period of significant expansion in U.S. space activity.
The methodology may be dated. The underlying question — how do you plan for failures that almost never happen, but matter enormously when they do — has not gone away.