ULA Launch Failure Delays Critical Military Satellite

On February 12, 2026, a United Launch Alliance Vulcan Centaur rocket lifted off from Cape Canaveral Space Force Station at 4:22 a.m. Eastern, carrying classified Space Force surveillance satellites toward geosynchronous orbit [1]. About 20 seconds after liftoff, one of the rocket's four Northrop Grumman GEM 63XL solid rocket boosters suffered what appeared to be a burn-through in its nozzle — a jet of flame shooting sideways from the booster casing [2]. The rocket's core-stage engines compensated, and the payload reached orbit. But the damage to the Pentagon's 2026 launch schedule had already begun.

The anomaly was not new. An identical failure had occurred during Vulcan's second flight in October 2024, when a nozzle separated from a GEM 63XL booster during the Certification-2 mission [3]. ULA and Northrop Grumman had attributed that incident to a manufacturing defect in internal insulators bonded to the nozzle and said corrective modifications were underway [4]. The February 2026 recurrence demonstrated that the fix had either failed or that a deeper problem remained unresolved.

Within two weeks, the Space Force made its response official: no more national security launches on Vulcan until the investigation is complete [5].

What Was on This Rocket — and What Comes Next

The USSF-87 payload included spacecraft from the Geosynchronous Space Situational Awareness Program (GSSAP), specifically GSSAP 7 and 8 [6]. These satellites operate in near-geosynchronous orbit — roughly 22,300 miles above the equator — carrying electro-optical sensors that track the behavior and movement of other satellites [7]. They serve as orbital sentinels for U.S. Strategic Command, monitoring whether foreign spacecraft are maneuvering near American assets. GSSAP provides a vantage point free from atmospheric interference that limits ground-based tracking systems [7].

The USSF-87 satellites did reach their intended orbit despite the booster anomaly [1]. The more consequential damage is to the missions still waiting on the ground.

Vulcan had been scheduled to fly seven national security missions in 2026 [8]. Among the most significant: the first Next-Generation Overhead Persistent Infrared (Next-Gen OPIR) satellite, a Lockheed Martin-built missile warning spacecraft that had already slipped from late 2025 to spring 2026 because of payload development delays and a crowded launch manifest [9][10]. Next-Gen OPIR is designed to replace the aging Space-Based Infrared System (SBIRS), which detects ballistic missile launches from geosynchronous orbit — a capability central to U.S. nuclear deterrence and missile defense [10]. The Vulcan grounding puts this launch in further jeopardy.

Three GPS III satellites originally contracted to fly on Vulcan have already been reassigned to SpaceX's Falcon 9 [11][12]. Space Systems Command stated the switch "ensures continued delivery of this critical system through responsive and reliable launch capabilities while the investigation into the Vulcan anomaly continues" [12].

The Technical Problem

The GEM 63XL is a solid rocket booster manufactured by Northrop Grumman at its facility in Promontory, Utah. Vulcan uses two or four of these strap-on boosters depending on mission requirements. Each booster generates roughly 449,000 pounds of thrust [4].

The recurring defect involves insulators bonded to the inside of the booster shell that forms the nozzle assembly [4]. During the October 2024 incident, investigators found that a manufacturing defect in one of these internal nozzle components caused it to detach, allowing hot exhaust gases to erode the nozzle structure [3]. A static test firing at Promontory in early 2025 was intended to validate modifications [13]. But the February 2026 recurrence — described by ULA as "a significant performance anomaly on one of the four solid rocket motors" — showed the same failure mode [2].

"This is going to be a many-months process as we work through the exact technical issue that happened and the corrective actions we need to take to make sure this doesn't happen again," Space Force Col. Eric Zarybnisky said at the Air Force Association's Warfare Symposium on February 25 [5].

ULA has stated it will not fly Vulcan again until the root cause is identified and corrective action implemented. The company and Northrop Grumman are recovering debris from the ocean as part of the investigation [4].

ULA's Reliability Record — and Its Erosion

For most of its existence, ULA built its reputation on a near-perfect launch record. The Atlas V, which Vulcan is replacing, maintained what ULA called a "100% mission success rate" over more than 15 years of service, with only two partial anomalies that did not prevent mission completion [14]. This track record was the company's central selling point to the Defense Department: ULA might cost more per launch, but the payload would get to orbit.

That argument has grown harder to sustain. ULA projected up to 20 launches in 2025 but completed only six — just one on Vulcan [15]. In May 2025 testimony before the House Armed Services Committee, Major General Stephen G. Purdy called ULA's recent performance "unsatisfactory," noting that slow progress in transitioning from Atlas and Delta vehicles to Vulcan had postponed four national security missions [16].

The Vulcan program itself has been marked by delays from the start. Originally slated for a first flight in 2020, the rocket did not launch until January 2024, after development problems with engines and an upper-stage explosion during testing at NASA's Marshall Space Flight Center [14][17].

Source: GDELT Project

Data as of Mar 21, 2026CSV

The Cost Question

The financial dynamics between ULA and SpaceX are stark. Under the most recent Phase 3 Lane 2 contract awards for fiscal year 2026, SpaceX received five national security missions valued at $714 million — roughly $143 million per launch. ULA received two missions valued at $428 million — approximately $214 million per launch [18][19]. ULA's per-mission cost is roughly 50% higher than SpaceX's for comparable national security orbits.

The longer-term Phase 3 contracts, awarded in April 2025, distribute approximately 54 missions from 2027 through 2032: SpaceX received a ceiling of $5.9 billion for 28 launches, ULA received $5.4 billion for 19 launches, and Blue Origin received $2.4 billion for seven missions [20][21]. The total portfolio represents nearly $14 billion in military launch spending.

ULA has historically received additional government support beyond per-launch payments. The Evolved Expendable Launch Vehicle program included "launch capability" payments — essentially retainer fees to keep ULA's infrastructure available. These payments totaled $983 million in fiscal year 2017 and $1.2 billion in fiscal year 2018 [22].

Public reporting has not identified specific financial penalties in ULA's contracts for launch delays resulting from technical anomalies. The Space Force has instead used its contractual authority to reassign missions — a practical penalty that reduces ULA's revenue and flight cadence.

One Provider Standing

With Vulcan grounded indefinitely, the Space Force's national security launch capability rests entirely on SpaceX's Falcon 9 and Falcon Heavy [5][8]. Blue Origin's New Glenn, while contracted for future NSSL missions, is not yet certified for national security payloads and is not expected to receive its first assignment until fiscal year 2027 at the earliest [20].

This single-provider dependency is precisely the scenario the National Security Space Launch program was designed to prevent. The Pentagon has maintained at least two "assured access to space" providers since 2005, when the policy was formalized to avoid the launch stand-downs that followed the Challenger and Columbia shuttle disasters [23].

Col. Zarybnisky indicated the Space Force is "considering a number of options" to keep missions on schedule [8]. The GPS III reassignments to Falcon 9 demonstrate one mechanism: contractual provisions that allow the government to swap launch providers when schedule or reliability concerns arise [11][12]. But transferring missions is not instantaneous. Each satellite must be integrated with its new launch vehicle, a process involving payload adapter modifications, acoustic and vibration analyses, and mission-specific trajectory planning.

SpaceX's Falcon 9 has launched more than 400 missions to date, with a success rate exceeding 99% [24]. But relying on a single rocket family — regardless of its track record — introduces concentration risk that military planners have long sought to avoid.

The Strategic Backdrop

The timing of the Vulcan grounding coincides with an acceleration in Chinese military space activity. China completed 73 orbital launches in 2025, up from 51 in 2024, deploying more than 300 spacecraft — roughly one launch every five days [25]. Chinese military programs include stealth satellite experimentation, on-orbit refueling demonstrations, and early deployment of the Three-Body Constellation, planned to eventually comprise 2,800 satellites [26][27].

Space Force officials disclosed in December 2025 that China and Russia have been experimenting with satellites designed to be harder to detect and track in orbit [26]. This is directly relevant to the GSSAP surveillance mission that Vulcan just delivered — the very satellites tasked with monitoring such activity.

Russia's military space program, while constrained by budget limitations, continued incremental expansion with GLONASS navigation satellite deployments and military-academy microsatellite launches through 2025 [28].

A January 2026 analysis in Defense News characterized 2026 as "a pivotal year for US readiness" in space, noting that delays to missile warning, space surveillance, and GPS refresh programs compound in ways that degrade the broader architecture [29]. Each satellite that reaches orbit late extends the operational life demands on aging predecessors — spacecraft already operating beyond their design lifetimes and with degraded capabilities.

ULA's Defense

ULA's argument has always centered on mission assurance — the principle that protecting a multi-billion-dollar national security payload justifies higher costs and more conservative timelines. The company's Atlas V record supports this position: across more than 100 flights, no payload was lost [14].

ULA has also pointed out that Vulcan's core stage and Centaur upper stage — the company's own hardware — performed nominally during the USSF-87 mission [1]. The booster anomaly involved Northrop Grumman's GEM 63XL, a component ULA procures from an outside supplier. ULA's press release after the launch emphasized that "the Vulcan booster and Centaur performed nominally and delivered the spacecraft directly to geosynchronous orbit" [1].

The company planned 18 to 22 Vulcan launches in 2026, with more than half for the Space Force — an ambitious ramp intended to demonstrate the vehicle's operational maturity [8][15]. That schedule is now effectively frozen.

ULA is also navigating a leadership transition. The company's interim leader has publicly committed to increasing launch rate and resolving outstanding technical issues [30]. Whether the organization can execute that plan while simultaneously investigating a recurring booster defect remains an open question.

What Happens Now

The investigation timeline is measured in months, not weeks. Each GEM 63XL booster in ULA's inventory will likely require inspection and potentially retrofit before flight [4][5]. Northrop Grumman must identify why its earlier corrective actions failed to prevent recurrence and demonstrate that new fixes are durable.

For the Space Force, the immediate task is triage: deciding which missions can shift to SpaceX, which must wait for Vulcan's return, and which payloads — like the heavy Next-Gen OPIR satellite — may be incompatible with Falcon 9's capacity to their intended orbits.

The broader policy question is whether the Pentagon's "assured access" model needs structural revision. With Atlas V and Delta IV retired or retiring, Vulcan was supposed to be the vehicle that carried ULA's share of the national security manifest for the next decade. Two identical booster failures in 16 months have put that plan on hold, leaving the military's most critical space capabilities dependent on a single company's rockets — the exact outcome the entire NSSL architecture was built to prevent.

Note: Several details of the USSF-87 payload and subsequent classified missions remain undisclosed at the request of the U.S. Space Force. The full scope of operational impact from the Vulcan grounding may not be publicly known.